News this Week

Science  14 Apr 2000:
Vol. 288, Issue 5464, pp. 238

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


    Florida Meeting Shows Perils, Promise of Dealing for Dinos

    1. Constance Holden

    Fort Lauderdale, FloridaIt was one of the most unusual coming-out parties Florida had ever seen. By day, some of North America's top dinosaur experts debated fine points of the evolution of birds and dinosaurs. After dark, they mingled with the cream of Fort Lauderdale society, supping, dancing, and drinking cocktails around a live alligator, assorted skeletons, and a dinosaur carved in ice.*

    The debutante was Bambiraptor feinbergi, a nearly complete specimen of an unusually birdlike dinosaur, which will make its home in Fort Lauderdale as Florida's first “real bone” dino fossil. The meeting was in part a tribute to the contributions of amateurs to paleontology. But it also held some cautionary notes about the problems that arise from the increasingly wild and woolly world of commercial fossil dealing: Conspicuously absent from the party was a key player in the not-so-happy story of another fossil—a doctored specimen called Archaeoraptor that was also on display at the meeting.

    Bambiraptor is the success story. Bambi was found in 1993 by the fossil-hunting Linster family at their private dinosaur plot in Montana. Aware of the significance of their find, they contacted professionals and eventually found a supporter in Michael Feinberg of Hollywood, Florida. Feinberg bought the fossil for a reported $600,000 and financed its reconstruction at the University of Kansas, Lawrence. Bambi is now the crown jewel of the newly created Florida Institute of Paleontology, the dinosaur branch of Fort Lauderdale's Graves Museum of Archaeology and Natural History.

    But the symposium, which drew 150 or so scientists, collectors, dealers, and grad students, also marked the end of a chapter in a less cheery saga: that of Archaeoraptor, a 124-million-year-old fossil found in the rich fossil beds of China's Liaoning province. Three days before the meeting, in Washington, D.C., scientists summoned by the National Geographic Society announced that Archaeoraptor is in fact a composite containing the remains of at least two animals—the body of a flying bird and the tail of a birdlike predator called a dromaeosaur.

    The Archaeoraptor story has been a major embarrassment for National Geographic, which last fall touted it as embodying “a dramatic combination” of bird and dromaeosaur characteristics. The glued-together slab of stone holding the fossil was purchased last year—reportedly for $80,000—at the Tucson Gem, Mineral, and Fossil Show by Stephen Czerkas, an artist who with his wife Sylvia operates a dinosaur museum in Blanding, Utah. Czerkas enlisted the help of paleontologists Philip Currie of the Royal Tyrrell Museum of Paleontology in Drumheller, Alberta, and Xu Xing of the Institute for Vertebrate Paleontology and Paleoanthropology in Beijing. He planned to publish a scientific paper about the animal, to coincide with a dinosaur story in the November 1999 issue of National Geographic. In August, Czerkas and Currie hired paleontologist Tim Rowe of the University of Texas, Austin, to perform a computed tomography (CT) scan of the fossil.

    Things started to go awry when both Science and Nature turned the paper down. National Geographic went ahead with publicizing the creature. But in December, Xu Xing, back in China, located a piece of rock containing most of a fossil dromaeosaur in possession of a farmer in Liaoning. Studying the tail, Xu says he became “100% sure” that it was the counterpart piece from the other half of the rock sandwich that held the tail of Archaeoraptor.

    On 4 April, Xu brought the fossil to Washington, D.C., where scientists meeting at the National Geographic Society examined the two specimens side by side and then issued a press statement saying that the specimen is “a composite of at least two different animals.” Rowe says his CT scans show that Archaeoraptor's legs may come from a third, unidentified animal.

    From a paleontological standpoint, Archaeoraptor has now split like stock—creating two new publishable specimens instead of one. Czerkas and Xu plan to co-author a paper about the bird part, which Czerkas presented at the Florida meeting as “a new toothed bird from China.” (Czerkas is still calling the bird Archaeoraptor, although Xu says it's time for a new name.) And Xu plans to write up the still-unnamed new dromaeosaur.

    Rowe and his CT scans, meanwhile, were nowhere near Fort Lauderdale. Rowe says he called the organizers of the symposium in late February, offering to present a talk on his results, but was turned down. One of the organizers, Wyoming paleontologist Robert Bakker, claims there wasn't room on the schedule to accommodate Rowe's “11th hour” request. Other sources, though, say Czerkas vetoed Rowe's participation. Rowe had recently submitted a paper on his results to Nature. In late March, Czerkas wrote to Nature and to the University of Texas threatening legal action if Nature publishes the paper without his permission. Rowe says the collaborative agreement he entered with 288Czerkas and Currie clearly allows him to publish his results. Czerkas, however, says of Rowe: “He was hired help. It's my right to publish first.”

    Rowe says the Archaeoraptor fiasco is typical of what can happen when paleontology and profits mix: “All the fossils that have come through commercial hands that I look at with my new eyes have been severely tampered with.” He says he warned Currie and Czerkas that preliminary scans indicated the specimen was “compromised” and might be a composite, but that Czerkas went ahead with the “publicity circus” at the National Geographic anyway. Czerkas responds that Rowe is operating from “20/20 hindsight.” He says, “We all agreed at the SVP meeting [the Society of Vertebrate Paleontology, which met in Denver in October] that the most parsimonious interpretation was the tail belonged.” It could not be determined from Rowe's CT scans whether the tail belonged with the bird body, Czerkas says; it was only Xu who came up with definitive evidence.

    Anyone who wants more evidence had better work fast; Czerkas agreed to return the Archaeoraptor composite fossil to China on 25 May.

    • *The Florida Symposium on Dinosaur Bird Evolution, 7 and 8 April.


    Dinos and Turkeys: Connected by DNA?

    1. Constance Holden

    They said it couldn't be done. But a team at the University of Alabama just may have succeeded in extracting some DNA from a dinosaur. And guess what it resembles: a turkey. If the work pans out, the scientists say, it will be the “first direct genetic evidence to indicate that birds represent the closest living relatives of the dinosaurs.”

    Last week at the Florida Symposium on Dinosaur Bird Evolution in Fort Lauderdale, biologist William Garstka of the University of Alabama in Huntsville reported that—with expertise from NASA and scientists from the Russian Academy of Sciences who have been probing for DNA in permafrost—he and colleagues may have isolated a stretch of mitochondrial DNA from 65-million-year-old Triceratops bones found in North Dakota.

    Because the bones were poorly mineralized, he says, the researchers think they were able to get a 130-base pair sequence from two vertebrae and a rib fragment. Matching the sequence against DNA samples from 28 animals, including 13 bird species, they found that it made a 100% match with the turkey and at least a 94.5% match with other birds. Naturally, says Garstka, “we thought of turkey sandwiches” that had probably been consumed both in the lab and in the field. But when the team checked for turkey DNA in turtle bones, dirt, and burlap from the same site, none tested positive.

    DNA from dinos is “for most people a truly heretical idea,” Garstka says, because many experts believe nucleic acid is unlikely to survive more than 100,000 years. Zoologist John Ruben of Oregon State University in Corvallis says, “The fact that turkey DNA was so similar to that of Triceratops was very suspicious.” Garstka himself says that “at this point, I remain quite skeptical of our own work. We would expect this kind of result from a theropod [a birdlike dinosaur], but here we're talking Triceratops.


    Stealth Genome Rocks Rice Researchers

    1. Elizabeth Pennisi*
    1. *With reporting by Dennis Normile in Tokyo, Pallava Bagla in India, and Li Hui in China.

    For the past 3 years, researchers from 10 countries, led by Japan, have been working on an ambitious effort to sequence the rice genome. Last week, many of the participants were stunned to learn that the biotech giant Monsanto is well ahead of them. Monsanto and collaborators at the University of Washington (UW), Seattle, announced on 3 April that they had almost completed a rough draft of the entire rice genome. Equally surprising, the company said it would turn its data over to the international consortium. “This is very big news,” said Takuji Sasaki, director of Japan's Rice Genome Research Program, with more than a touch of understatement.

    Monsanto's clandestine achievement is impressive. Not only is rice the first plant to be sequenced in rough form, but at 430 million bases it is also the largest genome ever sequenced—more than twice the size of the recently published Drosophilagenome (Science, 24 March, p. 2185). If the company lives up to its promise to make the sequence public, the International Rice Genome Sequencing Project could complete its work in just 2 or 3 years—and for half the estimated cost of $200 million. As a result, “public institutions committed to doing crop science research for developing countries' crops will have access [to the genome] much sooner, and without restrictions,” says microbiologist Gary Toenniessen, director of the Rockefeller Foundation's food security program.

    But because Monsanto (now a division of Pharmacia Corp.) kept the project secret—presumably to keep competitors in the dark while it got a first look at the sequence—few outsiders have seen the data, so it's hard to judge their quality or utility. And because there are few precedents for free public use of corporate data, some scientists are wondering whether the offer might be too good to be true.

    The rice sequence is the fruit of a collaboration between Monsanto and Leroy Hood, now president of the Institute for Systems Biology. The company gambled that a sequencing approach developed in part by Hood would quickly decode the rice genome. If so, the results could be useful for engineering rice, and they may also help in understanding corn and other crops in which Monsanto is interested. Hood's approach is a refinement of the strategy being used by the public-ly funded Human Genome Project. It entails fragmenting the genome, putting the pieces in bacterial artificial chromosomes for copying, and working out the nucleotide sequence of each BAC one at a time. The refinement developed by the Hood team, led by UW's Gregory Mahairas, is a “very efficient” way to determine which BACs to sequence, and in what order.

    The result, announced in simultaneous press briefings last week in Beijing and Tokyo, is a map that covers some 80% of the rice genome at least four times over—a good enough draft to enable gene prediction programs to find many of the estimated 30,000 genes, Mahairas says. Neither the UW team nor Monsanto would reveal how long the project took or how much it cost. Mahairas would only say that “the whole approach worked very, very rapidly.”

    Sasaki, who has seen the data, says the quality of the sequence varies from BAC to BAC, but “it's still very valuable.” Rod Wing, a molecular biologist at Clemson University in South Carolina who has been scrambling to determine the optimal set of rice BACs for the sequencing consortium, is more circumspect: “We're going to have to look at the data very closely” to determine how best to use both the public and Monsanto sets. Some partners may take up where Mahairas left off, using the Monsanto BACs directly in their sequencing efforts; Wing, on the other hand, is considering using the company's data to make the sequencing of his own BACs more efficient.

    Although some researchers wonder whether Monsanto will be as forthcoming with the data as promised, Hood insists it will. He describes the project as a “win-win situation”: Monsanto gets a jump start on finding the genes with commercial value, and the consortium saves several years and perhaps as much as $100 million.

    Rockefeller's Toenniessen holds out Monsanto's data-release policy as a model for other public-private collaborations. Some details are unclear, but as early as next month, the consortium will have access to much of Monsanto's data. Once a piece of Monsanto sequence goes into the consortium's public database, anyone—even competitors—can use it, no strings attached, says Sasaki. Until then, however, other academic researchers who want to use Monsanto's sequence must agree to give the company an option to negotiate nonexclusive rights to license any patents derived from its use. “It would be nice if other companies followed suit and made their fundamental genomics information available under similar circumstances,” says Toenniessen.

    DuPont, for example, has a private rice database, as does Novartis. Neither has released these data, but Novartis did help to launch the public rice effort by supporting Wing's research. Says Michael Bevan, a plant molecular geneticist at the John Innes Centre in Norwich, U.K., Monsanto's actions “certainly put other companies on the spot.”


    Hot Pepper Receptor Could Help Manage Pain

    1. Gretchen Vogel

    Bite into a hot pepper, and the pain that engulfs your tongue and palate really does feel like a burn. Several years ago, scientists uncovered the apparent reason: a cell-surface protein that in cultured cells responds both to heat and to capsaicin, the active ingredient in chili peppers. But cultured cells don't experience pain, and researchers weren't sure of the molecule's importance in animals. Now, genetically altered mice that possess an amazing tolerance for hot sauce have demonstrated that the protein plays a key role in several kinds of pain. The finding may eventually aid in the development of new pain-killing drugs.

    In work described on page 306, neuroscientists Michael Caterina, David Julius, and Allan Basbaum of the University of California, San Francisco, Martin Koltzenburg of the University of Würzburg, Germany, and their colleagues genetically altered mice to remove the receptor that responds to heat as well as to capsaicin and other so-called vanilloid compounds. The mice behaved normally in most respects, but showed less sensitivity to high temperatures and drank capsaicin-laced water freely. Their neurons also failed to respond to normally noxious stimuli.

    Those traits, Julius says, show that the receptor is not only “an essential part of vanilloid sensitivity” but also plays an important role in several other kinds of pain. Indeed, neurosurgeon James Campbell of The Johns Hopkins University School of Medicine in Baltimore says that the receptor is a promising drug target. “If we go after these receptors, we may be able to control [certain kinds of] pain,” he says.

    Researchers knew that neurons containing the capsaicin receptor, dubbed VR1 (for vanilloid receptor 1), respond to capsaicin and other painful stimuli in culture. But it was only after Julius and Caterina decided to inactivate or “knock out” the VR1 gene that they found that the resulting mice are impervious to capsaicin-induced pain.

    For example, capsaicin injected into the hind paw of a normal mouse causes the animal to lick and shake the tender paw. However, the mutant mice barely reacted to the injection, and their paws did not swell or become inflamed as much as they do in normal mice. When researchers laced the drinking water of normal mice with capsaicin, the normal mice took one sip, rubbed their snouts, and stayed clear of the water bottle. The mutant mice, however, drank happily, says Caterina, who is now at Johns Hopkins University.

    The mutant animals also tolerated high heat better, including having their tails immersed in a hot water bath and their paws put in contact with a hot plate. The animals did eventually react in both tests, showing that sensitivity was lessened, not eliminated, Caterina notes. This suggests that other heat-sensing channels play a role as well, he says.

    Another type of test suggests that VR1 plays a role in the extra sensitivity to heat usually displayed by inflamed tissues. Mustard oil painted onto the paws of normal mice causes them to become inflamed and very sensitive to heat—just as sunburned skin is seared by warm water or sunshine. But in the mice lacking VR1, the mustard-oil treatment did not enhance the response to heat, although the animals still displayed the hypersensitivity to touch that develops in inflamed tissues. Because touch-sensitive pain must be triggered by other neuronal responses, says Julius, the finding suggests that blocking VR1 would not relieve a common, painful condition—extreme sensitivity to touch, such as that accompanying shingles.

    The mouse work suggests, however, that such inhibitors may help combat another type of especially troubling pain, the chronic internal pain that can accompany tissue damage. Julius and his colleagues suspect that VR1 receptors might contribute to such pain. They found, for example, that neurons carrying the receptors can be excited by the acidic environment produced by inflammation. But neurons from VR1-deficient mice bathed in an acidic solution did not react as vigorously as neurons from normal mice did. Thus, the researchers hope that blocking the VR1 receptor might help relieve chronic internal pain.

    The fact that the VR1 knockout mice seem otherwise normal is encouraging for drug development, Campbell says. “It would appear that the [VR1 receptor] molecules are specific to pain-sensing neurons,” he says. That could lead to drugs with few side effects—perhaps only an inability to taste Tabasco sauce.


    Mir Gets New Lease on Its Scientific Life

    1. Andrew Lawler

    After more than 7 months in mothballs, the Mir space station is once again open for business. A Netherlands-based company called Mir Corp. is helping to fix up the aging and trouble-plagued facility to make it ready for researchers—and eventually rich tourists. But it could be a short-lived venture: The Russian government, which owns the 14-year-old facility, has not decided how long to keep it in orbit, and U.S. space officials say privately they would love to see it shut down once and for all.

    Mir Corp. has pledged to spend at least $20 million on the venture. It is backed by several wealthy American investors, including Washington, D.C.-based telecommunications millionaire Walter Anderson, and the majority shareholder is the Russian space company Energia, which operates Mir for the Russian government. As a start, Mir and Energia bankrolled the launch on 4 April of a Soyuz spacecraft carrying two cosmonauts to Mir. They will check out life-support systems on Mir's collection of pressurized modules, fix a small leak, and conduct some 50 Russian science and technology experiments. Mir Corp. officials hope that a successful mission will help convince Western governments and companies to reactivate experiments they already have on board and attract new paying customers.

    Jeffrey Manber, Mir Corp. president and a former Washington representative of Energia, argues that Mir offers opportunities “ranging from industrial production and scientific experimentation to space tourism and in-orbit advertising.” With completion of the international space station expected to slip from its scheduled 2004 date, Manber says Mir can serve as a temporary substitute for companies and governments that have set aside money for experiments: “There is already equipment on Mir which can be used very cost effectively.”

    Manber admits that doing science aboard Mir won't be a big moneymaker, and the company ultimately hopes to lure wealthy and adventurous tourists to visit the station. That idea received an unexpected boost last week from U.S. Transportation Secretary Rodney Slater, who applauded the company's efforts to create a space tourism business during a speech to aerospace industry officials in Colorado Springs, Colorado.

    But many U.S. aerospace industry officials, NASA managers, and others familiar with Mir are skeptical about the company's prospects. Mir has suffered computer and power shutdowns, a fire, and a collision with a resupply vehicle that damaged one of its modules. And Russia's financial troubles prevented significant upgrades during the 1990s. Moreover, the U.S. government in recent years has encouraged the Russians to deorbit Mir and concentrate the country's limited resources on building and launching its portion of the international station.

    “There's a tough road ahead,” says Manber, acknowledging the uncertain status of the facility. He is hopeful, however, that the current mission will be followed this fall by a crew conducting experiments on behalf of Western scientists. But for now, Mir is proving that space stations can have many lives.


    Claim and Counterclaim on the Human Genome

    1. Eliot Marshall

    J. Craig Venter stole the show last week. The day before Venter appeared at a hearing of a House science subcommittee on 6 April to review research on the human genome, his company, Celera Genomics Corp. of Rockville, Maryland, issued a press release announcing that it had “completed the sequencing phase of one person's genome.” The notice, which had a ring of finality about it, indicated that Celera's computers are poised to assemble the human data into a complete genome—a formidable task that Venter predicted at the hearing would take “3 to 6 weeks.” Sometime later this year, he says, he will make the data available on Celera's Web site. Celera's stock, which had fallen abruptly in mid-March, soared.

    It was an effective bit of propaganda: Celera released no new scientific data, but left the impression that it has bagged the human genome—just as it bagged the genome of the fruit fly in collaboration with the Berkeley Drosophila Genome Project earlier this year. But members of the nonprofit consortium that aims to complete a “draft” version of the human genome this year quickly tried to pour cold water on Celera's boast.

    Eric Lander, for example, director of one of the largest of the publicly funded sequencing centers, based at the Massachusetts Institute of Technology, advised reporters that a lot of work remains to be done. He was quoted in The Boston Globe as saying that Celera had only produced “a small fraction of the data required”—less, in fact, “than has been produced by the international public sequencing consortium.”

    A week earlier, the public consortium had indulged in some propaganda of its own. The National Human Genome Research Institute (NHGRI) announced that the nonprofit labs had sequenced the 2-billionth base pair of human DNA. As the genome is about 3 billion base pairs long, NHGRI director Francis Collins interpreted this to mean that the job was two-thirds done. Although the milestone is impressive, researchers say, it does not give an accurate reading of how near to completion the project is.

    Indeed, Venter went out of his way in testimony last week to downplay the consortium's achievements. “Mr. Chairman,” Venter said, “I find myself in the peculiar position of warning you that in the race to complete a draft human sequence, the publicly funded Human Genome Project may be at a stage where quality and scientific standards are sacrificed for credit. … Analysis of the public data in GenBank reveals that it is an unordered collection of over 500,000 fragments of average size 8000 base pairs. This means that the publicly funded program is nowhere close to being ‘done.’” Venter suggested that Congress urge the consortium's researchers to “keep their standards at the highest levels … and not rush to publish preliminary data for the sake of claiming priority.”

    Asked if there is any chance that the competing genome teams might still come together to finish this project, Venter said last week: “I keep trying to come to the dance, but the others are still taking lessons.” This prompted a member of the public consortium to respond: “We all want to go to the dance, but we can't agree on the music.” Given the harsh criticisms flying back and forth, collaboration seems unlikely.

    In fact, the competition could be moving to a new arena: Celera announced last week that it is immediately directing its army of 300 sequencing machines to analyze the genome of the mouse—which is widely seen as being critical for understanding the human genome. The public consortium began a mouse sequencing project late last year. Celera expects to finish its work on the mouse long before the public consortium, which is aiming to be done by 2005. But the consortium's mouse genome will be completed to fine detail and, unlike Celera's, it will be released on public Web sites.


    A Mold's Toxic Legacy Revisited

    1. Michael Hagmann

    In 1995, the Centers for Disease Control and Prevention (CDC) in Atlanta set off a cascade of alarms when an agency task force linked certain toxin-producing molds to a cluster of cases of sometimes fatal lung bleeding, or pulmonary hemorrhage, in infants. But last month, the CDC published the findings of two expert panels that identified what they called “serious shortcomings” in the initial investigation and concluded that “a possible association between acute pulmonary hemorrhage … and [mold] exposure … was not proven.”

    The reexamination is already stirring debate. Investigators involved in the original study are preparing a rebuttal of the CDC report to be posted on the Internet ( And resolving the issue is important, because sick infants may be just the tip of the iceberg of much broader public health problems. Toxic molds, which cause allergies and asthma attacks in sensitive individuals, have also been linked to the elusive sick building syndrome, which, in turn, has led to lawsuits and efforts to clean up mold-contaminated buildings—both costing millions of dollars.

    Dorr Dearborn, a pediatric pulmonologist at the Rainbow Babies' and Children's Hospital in Cleveland, triggered the original investigation in November 1994 when he alerted the CDC to a cluster of eight babies the hospital had treated for a normally rare bleeding of the lungs. The CDC immediately sent in a task force to look for possible causes. The team focused on the infants' homes. “We realized that it must have to do with a home environment problem,” Dearborn recalls, “because if we sent the infants home again, they restarted bleeding.”

    It turned out that all the houses with sick babies had recent water damage. After sampling these homes and several control houses in the same area, the investigators concluded that the likely culprit was Stachybotrys chartarum and other toxic molds that thrive in damp buildings and can under certain conditions produce spores containing a nasty cocktail of toxic chemicals. Although the investigators cautioned that more research was needed to prove the case, their findings precipitated a frenzy of activity. Public health guidelines were issued, contaminated buildings were evacuated and closed, multimillion-dollar lawsuits ensued—and the media jumped on the bandwagon.

    But many in the scientific community felt that some questions remained unresolved. So in November 1997, then-CDC director David Satcher asked an internal working group and a panel of outside experts to review the Cleveland investigation. The groups delivered their reports last June and December, respectively, and the CDC published a synopsis in the 10 March issue of the Morbidity and Mortality Weekly Report (MMWR) (

    Both panels spotted several flaws in the Cleveland study. For example, investigators collected twice as many samples in sick infants' homes as in control homes, and did so much more rigorously, the report states. “It's no surprise if you find more fungi in case homes this way,” says Brian Shelton, a microbiologist at Pathcon, a private laboratory that specializes in building and environmental health assessments.

    There was also no clear-cut clinical definition of the so-called idiopathic pulmonary hemorrhage. “The mere presence of blood, it seems, was enough to include infants as cases,” says expert panel member Alan Cohen of Georgia Pediatric Pulmonology Associates, an Atlanta-based private association of pulmonologists. “But how can you define a common cause if you don't even have a defined disease?” And a statistical reanalysis of the original data indicated that the results might have been skewed by the finding of “extremely high, outlying values” for S. chartarum contamination of one home. This “magnified the risk about fivefold,” says Daniel Sudakin, a medical toxicologist at the Veterans Administration Medical Center in Portland, Oregon.

    These and additional minor problems, taken together with other evidence from the literature, led the panels to conclude that S. chartarum's role in pulmonary hemorrhage was not proven. “That doesn't mean that S. chartarum is dismissed as a possible cause, but right now we just don't know what killed the Cleveland babies,” says Cohen.

    Dearborn acknowledges that because their study was designed rapidly, “it can't be perfect.” But, he says, the “minor deficiencies are not enough to invalidate our conclusions.” As support, Dearborn cites the fact that the number of infants with pulmonary hemorrhage has gone down recently in Cleveland—a change he attributes to public health officials inspecting homes for water damage and mold, and then having any contamination cleaned up.

    Both sides do agree on one thing: Further studies are needed. “We could be missing something that is right in front of us because we think we already have the answer,” Cohen says. And despite his disagreement with the MMWR report, Dearborn is happy the CDC is again studying the topic. “While we have continued our research efforts, the CDC stopped surveying and looking at pulmonary hemorrhage a few years ago. Now they're willing to do follow-up studies—that's great,” he says.


    Pruned Sanctions List Points to Closer Ties

    1. Jeffrey Mervis*
    1. * With reporting by Pallava Bagla in Hyderabad.

    For 2 years, a $500,000 scintillation counter built by Indian scientists has been sitting unused at the Fermi National Accelerator Laboratory (Fermilab) in Batavia, Illinois. The reason: U.S. sanctions, imposed after India's latest nuclear weapons tests, have prevented the team that built the device from coming to the United States to test and install it. But last week, in a development seen as a harbinger of greater cooperation between the two countries, three Indian scientists finally arrived at Fermilab to work on the equipment.

    After India's May 1998 tests, the U.S. government prohibited interactions between U.S. researchers and scientists at some 200 Indian institutions deemed to be part of the country's nuclear weapons and missile establishment. But on 18 March—the day before President Clinton visited India, where he declared that the bilateral relationship “was too important to ever fall into disrepair again”—the government formally removed 50 institutes from this so-called “entities list.” Among them was the Tata Institute for Fundamental Research in Mumbai, which built the scintillation counter. Just 3 weeks later, three researchers from the institute arrived at Fermilab to work on the device, which is part of a massive detector called D0 (D-Zero), for Fermilab's Tevatron accelerator. Seven other Tata researchers are expected to follow in the next 6 months.

    The easing of sanctions is part of a broader U.S. attempt to find areas of cooperation despite India's refusal to sign the Comprehensive Test Ban Treaty and abide by U.S. rules aimed at preventing the spread of missile and other “dual use” technologies, say U.S. officials. “It's a dual message,” says one State Department official. “We acknowledge that there are differences [between the United States and India], but we say that it's time to move forward.”

    Indian science officials welcome the move, but say they are disappointed that an all-day “roundtable” meeting during Clinton's visit didn't open up more civilian research to joint activities. “Cooperation would be furthered if the roundtable had come up with a better and more constructive definition of dual use,” says Kota Narayanan, director of the Aeronautical Development Establishment, a defense institute that remains on the banned list. “Given the imagination, anything can be classified as dual use.”

    For physicist Naba Mondal, the first of the Tata team to arrive at Fermilab last week, the easing of the sanctions on Tata researchers came none too soon. “It's a relief to participate again,” says Mondal, who last worked at Fermilab in early 1998. His Fermilab colleagues say they are glad to have him. “We have enough other problems to solve, so it's great to have them here to work on their instruments,” says Harry Weerts, a physicist at Michigan State University in East Lansing and a spokesperson for the D0 team. The detector is scheduled to be up and running in March 2001, 9 months behind schedule but still in time for the Tevatron's next set of experiments.


    Cancer Drugs Found to Work in New Way

    1. Marcia Barinaga

    When it comes to treating cancer with drugs, the dogma is “no pain, no gain.” Patients are hit with doses that may take them within an inch of their lives, then allowed to recover for several weeks before being blasted again. Occasionally, when patients can't tolerate or have already failed to respond to high-dose chemotherapy, oncologists try a gentler chemotherapy regimen: low oral doses taken continuously. Although this approach minimizes side effects and sometimes even shrinks tumors, it does not work well enough to be widely used. Nor has anyone understood its mode of action, especially when it slows down tumors that have already developed resistance to the drug. Now answers are emerging from two new studies in mice, and the information may enable clinicians to improve the effectiveness of this type of therapy.

    The studies come from Timothy Browder, Judah Folkman, and colleagues at Harvard Medical School in Boston, who describe their results in the 1 April issue of Cancer Research, and from another team, led by Giannoula Klement and Robert Kerbel of the University of Toronto, who presented their findings last week at the annual meeting of the American Association for Cancer Research (AACR) in San Francisco. (Most of what they reported appears in the 15 April issue of the Journal of Clinical Investigation.) Both teams show that this gentler form of chemotherapy, recently dubbed “metronomic” therapy because it never misses a beat, may work by blocking angiogenesis—the sprouting of new blood vessels that feed growing tumors. What's more, both groups show that the effectiveness of metronomic therapy is enhanced when it is used in combination with drugs that specifically inhibit angiogenesis.

    Folkman and Browder set out 5 years ago to find out why standard chemotherapy, which kills dividing cells, doesn't block angiogenesis by killing the endothelial cells that divide to form new blood vessels. If chemotherapy did target those endothelial cells, then it should kill even drug-resistant tumors by blocking their blood supply. Working in mice with tumors, Browder figured out why that doesn't happen: With standard intermittent chemotherapy, the endothelial cells recover during the rest periods and restore the tumor's blood supply.

    In the current work, both the Harvard and Toronto teams report that if they eliminate the rest periods, they can prevent this from happening. The two teams inoculated mice with tumors, including some that were highly resistant to the chemotherapy drugs they were using—cyclophosphamide in the Harvard group's experiments, and vinblastine in the Toronto group's. Both found that continuous treatment with relatively low and easily tolerated doses of the drugs caused the tumors to shrink or slowed their growth. “That suggests we are having an effect on some other [nontumor] cell type,” says Kerbel. Indeed, Browder's work confirmed that the treatment kills endothelial cells and blocks angiogenesis.

    The tumors eventually regrew. But when the teams added a known antiangiogenic drug to the mix—the Harvard team used a drug called TNP-470, and the Canadians used an antibody that blocks the receptor through which vascular endothelial growth factor, VEGF, exerts its effects—the tumors did not return, even when treatment was discontinued.

    Not only does this suggest that the treatments cured the mice, says Kerbel, but his team saw “no overt toxicity” in the treated animals. The Harvard mice also fared well, losing only 5% of their body weight—compared to 20% on standard chemotherapy—and living out their full life-spans. “I'm very excited” about the promise of combining antiangiogenesis drugs with metronomic therapy, says cancer biologist Douglas Hanahan of the University of California, San Francisco, who wrote a commentary on the Kerbel team's paper. “There could be some real benefits there.”

    The extent of benefit for humans remains to be seen, but some cancer researchers have been sufficiently optimistic to undertake clinical trials. A team at the European Institute of Oncology in Milan has begun a study of metronomic chemotherapy, using the drugs Cytoxan and methotrexate, in patients with breast or colon cancer. At the AACR meeting, team member Filippo de Braud reported that some patients are showing tumor shrinkage. He also said the team plans to combine the therapy with an antiangiogenesis drug.

    But Kerbel and Folkman caution that oncologists should not put patients who have other options on metronomic therapy unless clinical trials prove it to be effective in humans. Even if clinical studies do show benefit, they expect the approach will be less successful in humans than in mice, given experience with other cancer drugs. But, Kerbel adds, if the work leads to a new cancer therapy that “prolongs survival in a subset of cancer patients, with minimal or no toxicity, that will be a very significant advance.”


    Transgenic Crops Report Fuels Debate

    1. Jocelyn Kaiser

    Wading into one of today's most politically charged scientific issues, a National Academy of Sciences panel* last week called for tightening the regulation of plants genetically modified to repel pests. Transgenic crops have generally been adequately tested for health and environmental effects, but agencies should collect more data and coordinate their reviews, concluded the panel. In keeping with the drama that accompanies anything about genetically modified organisms (GMOs), industry groups immediately trumpeted the report's conclusion that biotech foods on the market are safe, while environmentalists dismissed the report as “tainted” by industry ties.

    The long-awaited study is the first academy report in more than 10 years on biotech crops, which are flooding the market. Indeed, more than one-fifth of all corn and cotton crops planted in the United States last year contained a bacterial gene for a pest-killing toxin called Bt. Many activists and some scientists have argued that the health and ecological risks of these plants haven't been adequately assessed (Science, 26 November 1999, p. 1662). On the flip side, a number of scientists have voiced concerns about overregulation. A coalition of 11 scientific societies has been lobbying the Environmental Protection Agency (EPA) to scrap a 1994 proposed rule that regulates transgenic “pesticidal plants,” arguing that it is unscientific to regulate the process, genetic engineering, as that could encompass features as innocuous as pest-repelling hairs on a plant's leaves (Science, 9 April 1999, p. 249). Instead, the societies argued that EPA should regulate the plant's products, such as expressed proteins that might be toxic.

    The academy panel, chaired by Perry Adkisson, an entomologist and chancellor emeritus at Texas A&M University in College Station, was formed a year ago partly to address scientists' concerns about the EPA rule. Looking only at what it termed “transgenic pest-protected plants,” the panel endorsed their use, saying they could help to reduce the amount of chemical insecticides applied. The panel also dismissed health concerns: “The committee is not aware of any evidence that foods on the market today are unsafe to eat as a result of genetic modification.” But it urged more research on, for instance, the flow of genes from crops to weedy relatives, long-term ecological effects of transgenic crops, and potential health effects, monitored through long-term animal feeding studies.

    As for EPA's proposed regulations, the panel came down firmly on the side of keeping—indeed strengthening—them. It recommended scrapping two EPA exemptions that assume certain plants are safe: those made by adding viral coat proteins (because the virus could spread to weeds), and those made by inserting a gene from a plant similar enough to interbreed. And it suggested that regulatory agencies add a few requirements—for example, tests for protein allergenicity—and share their data with the public.

    The panel's report is “schizophrenic,” says R. James Cook, a plant scientist at Washington State University in Pullman and spokesperson for the 11 scientific societies. Cook wonders why the panel endorses a different type of regulation for transgenic crops while concluding that they are not inherently more risky than traditional crops. The answer is simple and pragmatic, says panelist Fred Gould, an entomologist at North Carolina State University in Raleigh: “If you got rid of that rule, public confidence would be down the toilet.”

    Even so, public confidence could still use some shoring up. Although the Biotechnology Industry Organization (BIO) was delighted with the report—it issued a press release proclaiming that transgenic foods “are thoroughly tested and safe”—many activists weren't. Before the report was released, protesters gathered in front of the academy with Representative Dennis Kucinich (D-OH). He urged the academy to “scrap the study” because the panel was “tainted by pervasive conflicts of interest,” including the departure of the study's original director, Michael Phillips, last July for a job with BIO. The academy concedes that two panel members—an attorney and an industry consultant—did have conflicts of interest, but, according to executive officer William Colglazier, “we felt their regulatory expertise was needed.” An internal investigation determined that the report was not biased by Phillips's involvement, he says. The one activist on the panel, ecologist Rebecca Goldburg of Environmental Defense, concurs. “Obviously, I think the panel had enough to offer that I stuck with it.”


    'Pre-Clovis' Site Fights for Recognition

    1. Erik Stokstad

    One of the fiercest battles in paleoanthropology concerns the peopling of the Americas: Were the first Americans so-called Clovis hunters who crossed from Asia about 12,000 years ago, or did others get here first, perhaps crossing the Pacific or the Atlantic? At the annual meeting of the Society for American Archaeology, held last week in Philadelphia, a team of researchers presented evidence that humans camped many times on a site in Virginia dated to 18,000 years ago. Distinctive stone tools, found at a site called Cactus Hill, lie below artifacts typical of the Clovis people, who take their name from an 11,500-year-old site in Clovis, New Mexico. Many researchers are wary of the dates, but others say the results are a strike against the Clovis-first theory. “It's step one of accepting it as pre-Clovis,” says Dennis Stanford of the Smithsonian Institution's National Museum of Natural History.

    How old?

    Humans may have camped here at Cactus Hill as many as 18,000 years ago.


    At one level in the Virginia site—a large, sandy hill some 70 kilometers south of Richmond—the team found classic Clovis blades dated to about 10,000 years ago, says Joseph McAvoy of Nottoway River Survey, a private archaeological consulting firm. Some 15 centimeters below, they uncovered subtriangular projectile points that were clearly not like Clovis artifacts. Radiocarbon dates from associated charcoal suggested that the tools were 5000 years older than the Clovis points above, as documented in an extensive 1997 report. But many archaeologists worried that the unusual artifacts might have fallen from above and been mixed in the sand by plant roots or burrowing animals.

    McAvoy marshaled an interdisciplinary team of 15 researchers to find out. Lucinda McWeeney of Yale University's Peabody Museum of Natural History found a much higher concentration of silica remains from plants at tool-bearing levels. She says that's consistent with the idea that people were camping and bringing in plant materials, although others point out that this may be due to ecological succession. She also noticed matching peaks of phosphates, perhaps from urine or excrement.

    Meanwhile, Daniel Bush and James Feathers of the University of Washington, Seattle, dated the sand samples with a process called optically stimulated luminescence, which measures the time elapsed since grains were exposed to light. This backed up the radiocarbon dates. Moreover, five samples showed virtually no vertical mixing of grains, McAvoy adds. “There's still some uncertainty about the absolute age of the pre-Clovis deposits, but the relative sequence looks very good,” says David Meltzer of Southern Methodist University in Dallas.

    Meltzer and others caution that the artifacts may not be the same age as nearby 18,000-year-old charcoal, because particles are often transported inside sand deposits. The team did get a wide range of radiocarbon dates—both pre- and post-Clovis—on charcoal associated with the pre-Clovis tools. McAvoy's team dismisses the younger dates as due to contamination. But others aren't so sure. “If there's no consistent pattern [in the dates], then there may be a problem with mixing” of charcoal from different levels, Meltzer says.

    Still, the fact that the tools are roughly similar to those of another possible pre-Clovis site in Pennsylvania, called Meadowcroft, is good news. “These are not isolated things that we can't make sense of,” Meltzer says. “The point forms bear some resemblance to each other. We're starting to see commonalties, and that's heartening.”


    The Rise of the Mouse, Biomedicine's Model Mammal

    1. David Malakoff

    Molecular geneticists looking for ways to model human disease and companies testing new drugs are creating an unprecedented demand for inbred rodents

    Bar Harbor, MaineIn a shiny new $23 million facility here at The Jackson Laboratory (TJL), mouse pups rate their own private elevator. The miniature “mousevator” whisks the thumb-sized newborns from ground-floor surgical suites up to germ-free nurseries, where whiskered foster parents and a bevy of human coddlers await. These animals—bred to mimic human diseases at costs approaching thousands of dollars each—are too valuable to be left to nature's mercies, notes geneticist Larry Mobraaten, a driving force behind the new Genetic Resources Building, dedicated to raising new and useful mutants. “They are extraordinarily valuable scientific resources,” he says.

    The mousevator is just one indication of the booming mouse economy. From lab benches to Wall Street, everyone from venture capitalists to cagemakers is scurrying to get in on the unprecedented international growth in the use and production of mice for scientific research. By some estimates, more than 25 million of the tiny mammals will be raised worldwide for studies this year, accounting for more than 90% of all mammals used in research. That's double the number used a decade ago, but still not enough to meet anticipated demand: Forecasters predict mouse use could grow by 10% to 20% annually over the next decade. “It's a feeding frenzy,” says Ken Paigen, director of TJL, which ships nearly 2 million of its trademarked JAX mice to researchers each year.

    Once a modest regional business, the mouse trade is now a global enterprise that is being transformed by scientists' growing ability to fine-tune the genetic variation of these model mammals. Major commercial breeders sold an estimated $200 million worth of rodents last year; now they are retooling to meet greater demand spurred by government and corporate spending on biomedical research. Universities are also aggressively building new animal facilities to lure top scientific talent, while lawyers on the “mouse bar” struggle to settle lingering controversies surrounding patented mice (see pp. 254 and 255). And government agencies are pouring cash into programs that will produce a new network of distribution centers and an avalanche of new strains—raising questions about which ones should be maintained as live breeding colonies and which ones frozen in vaults for future use. Meanwhile, researchers and breeders are keeping a nervous eye on animal rights activism, regulatory initiatives, and mouse diseases that could complicate these plans.

    Several factors have contributed to the boom, including the mouse's spectacular fecundity and relatively low maintenance costs. Some prolific pairs, for instance, can produce more than 250 descendants in just a year on little more than grain and water. But scientists like mice because they are physiologically and genetically similar to humans. Millions of mice are used to screen drugs and potentially dangerous compounds for safety, for instance. And most human genes appear to have a related mouse version, making it possible to gain insights into human diseases using gene-altered mouse models that suffer from similar ills but aren't subject to the same ethical concerns as human patients. Technologies that have made it easier than ever to tinker with the mouse's genome have only enhanced the rodent's value. For instance, the potential number of transgenic and “knockout” mice (which have one or more of their 80,000 genes disarmed) is mind-boggling, notes Donna Gulezian, product manager for transgenic models at Taconic Farms, a major mouse supplier in Germantown, New York. Mouse design, she says, is “limited only by the investigator's imagination.”

    Indeed, the mouse's growing importance as a “fuzzy test tube” and its close kinship to humans has made it the only other mammal scheduled for complete genetic sequencing, a task that both the National Institutes of Health (NIH) and the private company Celera have targeted to complete within 5 years. The honor is one sign of the rodent's transformation from “lab urchin to scientific thoroughbred,” says Bob Jacoby, director of the Animal Resources Center at Yale University in New Haven, Connecticut.

    The mousketeers

    Like many businesses, the modern mouse economy is dominated by a few big names that coexist with some well-respected niche players and cottage industrialists. Globally, the commercial breeding heavyweight is Charles River Laboratories of Wilmington, Massachusetts, a 50-year-old concern with 49 facilities in 18 nations. Last year, it sold more than $140 million worth of mice, rats, and other research animals. (Company officials declined to detail how many mice they sold.) The other two U.S.-based, multinational industry leaders—number two breeder, Harlan Sprague Dawley of Indianapolis, Indiana, and Taconic—are smaller. Analysts estimate that Harlan had $60 million in 1998 sales, while Taconic totaled $36 million last year. Also in the top pack is TJL, which rang up $29 million in mouse sales in 1999. But unlike its competitors, the lab—which one executive calls “the fourth mouseketeer”—is a taxpayer-funded nonprofit that plows profits back into research and warehousing thousands of mouse varieties that have little commercial value.

    Charles River and Harlan are also major players in Europe, having purchased stakes in a number of homegrown providers. They are joined by Taconic ally M&B of Ry, Denmark—a significant supplier in Northern Europe and Germany—and government-sponsored mouse repositories, such as the European Mutant Mouse Archive in Italy. Charles River also has outposts in Japan, where it competes with CLEA Inc. of Tokyo and smaller suppliers.

    On a regional scale, academic research labs also sustain a lively barter in specialized mice among scientists. Also serving a limited clientele, but charging big fees, are a few specialized high-end companies such as Lexicon Genetics of The Woodlands, Texas. For prices ranging from $18,000 to $65,000, depending on the company's share in any royalty income, Lexicon will spend 8 months creating four custom-tailored knockout mice for each customer.

    The bulk of the world's lab mice, however, are bred by large academic or government labs for internal use or supplied by the high-volume breeders. And most of these suppliers are now reengineering themselves to keep pace with growing demand. Charles River, for instance, is completing a major financial reorganization after being sold last year by corporate parent Bausch & Lomb. The buyers—who paid $456 million—are Charles River's own longtime executives, backed by Global Health Care Partners, an investor group that includes the former CEOs of several major pharmaceutical companies.

    These executives are making what one industry insider calls a “gutsy gamble.” Other breeders are impressed with the amount of debt that Charles River has taken on—nearly $350 million—given the risks of the live-animal trade, from mergers that can trim customer lists to diseases that can wipe out a close-packed breeding colony virtually overnight. Still, documents filed with the Securities and Exchange Commission in Washington, D.C., show that the company is the high-volume Wal-Mart of the mouse economy, with 62% of its $230 million in sales from animal models. Company executives are bullish that they can build on this position, noting that Charles River historically has “been able to increase our prices at rates that are above the rate of inflation … by maintaining high quality.” They predict “moderate but sustained growth in the research model business.”

    Harlan, Taconic, and TJL are also planning for multinational growth. TJL, for instance, is adding capacity. For the first time, it has also hired a business-savvy executive solely to manage and grow its production division, which generates nearly half of the funds the lab pumps into its research programs. New head mouse wrangler Warren Cook, a veteran of chemical and skiing businesses, says his top priority is improving the lab's ability to deliver mice in a timely manner, long a sore point with some researchers.

    The problem is that while most suppliers offer 25 to 75 strains, TJL has about 2500—“by far the world's best selection,” gushes one longtime customer. But diversity is also the lab's Achilles' heel, she says: “They can't keep every strain on hand, so you sometimes have to wait a long time for delivery.” Cook jokes that buyers put up with delays because, “despite our lousy service, we still know mice better than anyone.” But increasingly, says TJL's Phil Standel, the wait is unacceptable. Customers, he says, “see mice just the way you see a chemical kit you can order off the shelf in a few days.”

    Like TJL, Harlan and Taconic have added customer-service staff. Following an industrywide trend, they are also offering customized breeding services to clients who want to avoid the high cost of housing or raising their own mice, particularly the hard-to-maintain transgenic types. Taconic technicians, for instance, now care for more than 600 lines of “outsourced” mice belonging to other labs, along with about 75 of their own strains.

    Contract breeding is attractive not only because it generates income. It also can give breeders an inside track on emerging models that may be worth adding to the product line. Indeed, to a greater degree than its competitors, Taconic is specializing in breeding the temperamental transgenics. It has created a “Transgenic Exchange” that helps researchers share their not-quite-ready-for-prime-time mice with other scientists. In exchange for Taconic's help in distributing and developing the model—a complicated breeding and characterizing process that can take years—the company positions itself to add the more popular contract-bred mice to its glossy catalog. Transgenics already account for nearly 10% of Taconic's revenue and are expected to be “a big part of our future,” says part-owner and director of marketing Sam Phelan.

    But few researchers should count on a cushy retirement as a mouse tycoon, industry officials caution. “We get calls all the time from researchers who think they've created the next big mouse,” says Taconic's Gulezian. “But the reality is very few models have broad enough applications to be commercially attractive.” At TJL, for instance, “most of our strains are money-losers,” but “it serves our public purpose to maintain them,” notes Paigen.

    There are exceptions. Although details are shielded by proprietary agreements, researchers who invented now-widespread patented techniques for engineering mice or who hold stock in biotech companies with rights to unusually useful strains have done very well. Earlier this year, for instance, a mouse engineered to grow human tissues proved so valuable that it scuttled a planned $350 million buy-out of a California biotech company. The promise of the Xeno mouse, owned by Abgenix Co. of Alameda, so excited investors that the company's stock value shot from $130 million to $370 million in just a few weeks—making the buy-out offer pale in comparison.

    But the big mouse breeders can't count on controlling such patented mice; instead they rely on their brand name to market common strains available from many vendors. TJL, for instance, bans direct sales to its competitors, in order to “maintain the strength and integrity of our brand,” says Cook: “Our Black 6 is different from Charles River's Black 6.” And TJL's Carol Linder adds that studies have shown that mice from different vendors have developed significant genetic differences over time, though they may share the same name. “I'd never recommend switching suppliers midway through an experiment, even if you think you are ordering the same mouse,” she says.

    Down on the ranch

    The toughest part of mouse ranching, however, may not be differentiating your product but keeping it healthy. “Raising mice can be a nightmare,” says Phelan. “Most researchers are blown away when they see what it takes to run a production facility.” Companies spend millions, for instance, to prevent human caretakers from infecting their wards with disease. At Taconic and elsewhere, masked and gowned workers are required to shower and don sterile jumpsuits before entering “barrier facilities”—mouse barns with sophisticated ventilation and watering systems. Some of the immunocompromised transgenic and mutant mice are particularly vulnerable and must be housed in germ-free plastic bubbles. (To introduce the “good” microbes mice need to digest food, caretakers often add a single pellet of mouse feces to their drinking water.) Other strains can't stand bright light, need cages mounted on vibration-damping shock absorbers, or stop reproducing or die if their food or ventilation isn't just right. Abigail Smith, an animal-care specialist who recently left Loyola University in Chicago, Illinois, for TJL, recalls that one strain would “start seizuring if you just clapped your hands.”

    Despite the precautions, almost every producer has had to destroy vast numbers of animals to halt epidemics. Indeed, disease is such a grave threat to sales that major producers are quick to investigate and address any suggestion that their animals are contributing to an outbreak. As rumors spread last year that TJL and Taconic mice appeared to be testing positive for a feared mouse hepatitis virus, for instance, both companies took aggressive steps to clear their names. After detailed testing—the results of which were posted on their Web sites—Taconic and TJL researchers concluded that the “outbreak” was either a rash of false-positive results or a hepatitis strain spread by mice from some other source.

    Mouse suppliers obsess over animal health in part because studies have shown that mice carrying pathogens can produce flawed research results. However, suppliers—and researchers—are also becoming sensitive about the high price tags on some mice. “When a mouse cost a buck and it got sick, no problem: You'd get another one,” notes Yale's Jacoby. But with transgenic mice routinely costing $175 each, and some rare pairs worth up to $30,000, “providing health care is becoming an increasingly attractive option,” he says. Mouse doctors—and pathologists, for essential postmortems—are in short supply, however. As a result, the NIH is calling for training a new generation of specialists who can keep animals healthy and help researchers understand the sometimes subtle genetic and environmental factors that influence an animal's behavior and physiology.

    Disease concerns have also prompted renewed calls recently for international testing standards; researchers want to know that the mice they get are clean. Several decades ago, animal-care experts thought they had solved that problem by introducing “specific pathogen free” (SPF) standards. The push, which prompted mouse users to take greater care in testing and accepting new mice into their colonies, helped produce a dramatic leap in health quality. Before SPF, “your average mouse was basically a sewer—it had every microbe known to mousedom,” recalls Smith, adding, “Things are much better now.” Today, however, the SPF label is used so routinely and enforcement is so lax that it has become virtually meaningless, some animal-care experts say. Health enforcers have not kept up with the proliferation of new and newly detectable mouse diseases, says Smith, who calls SPF “a garbage term unless they specifically tell you what pathogens they've tested for.”

    To restore SPF's good name, some countries, such as the United Kingdom, have adopted new policies that prohibit laboratories from accepting animals that haven't been certified as free of a “hot list” of pathogens. So far, however, major producers in the United States, Japan, and elsewhere in Europe have resisted the regulations, arguing that they are unnecessary in a market that already places a premium on health. Says Taconic animal-care chief James Geistfeld: “Our approach is to test the heck out of the animals and then publish the results.”

    Quiet resistance to greater health regulation also comes from researchers impatient to begin experiments with newly acquired animals. Some scientists sneak untested animals around quarantine restrictions, mouse health experts claim. The results can be disastrous. Loyola, for instance, had to shut down its colonies earlier this year after a pathogen was introduced by what officials believe were smuggled-in animals, disrupting dozens of experiments. It can take a year or more to complete the expensive—about $5000 per strain—process of rederiving stocks by implanting embryos in disease-free foster mothers. But Smith believes that, as scientists become more aware of the risks of working with untested animals, “they'll respond appropriately”—perhaps by turning in rogue colleagues.

    Companies not only need a clean bill of health, but they are increasingly pressured to certify that their mice are—genetically speaking—the real thing. Many researchers have horror stories about mice that turned out to be genetically different from the advertised strain. Even these problems, however, can be an opportunity to mouse providers: For a fee, they will do the sophisticated genetic testing necessary to cull imposters.

    Hickory, dickory, stock

    As mouse strains proliferate, one of the biggest challenges facing every retailer is figuring out how to keep live mice “on the shelf” awaiting a buyer. “One of the hardest things to control is fluctuating demand,” says Taconic's Gulezian. The problem is especially acute at TJL, which as a federally funded mouse repository has a mandate to keep as many potentially useful strains on hand as possible. Deciding what to keep is becoming more difficult: Scientists will create more strains this year than used to be developed in an entire decade. And the problem will only get worse, as special mouse initiatives in Europe and the United States ramp up.

    The U.S. initiative—championed by former NIH director Harold Varmus—includes millions of dollars to create potentially thousands of new mutants and transgenic mice. Last December, for instance, the National Cancer Institute funded 19 groups at 30 institutions to “accelerate the tempo at which mouse models of cancer are developed.” And NIH officials have tapped TJL, Taconic, and Harlan to help operate a new network of regional distribution centers that will help house and characterize new mutants created by exposing mouse sperm to ethylnitrosourea, a powerful mutagenic chemical. A related effort by the European Community hopes to pump up stocks at the European Mutant Mouse Archive. Although such centers will help ease the housing shortage, selection panels will still face some tough choices. “We'll have to do some crystal balling about what will be in demand years from now,” says TJL's Mobraaten, whose facility can accept about 90 new mutants a year.

    In the long run, however, it will be impossible for mouse researchers to build their way out of the space shortage, observers say. “You can't throw bricks, mortar, plastic, and stainless steel at the problem forever,” says Yale's Jacoby. Like others, he is hoping that new storage technologies—from freezing embryos or eggs to sperm and chunks of ovary—will eventually reduce the need to maintain live colonies. With that in mind, Mobraaten can equip TJL's new building with up to 18 cryogenic freezers—up from an existing four. But he notes that, so far, only the relatively expensive embryo-freezing process has proven effective with mice, while newer sperm- and ovary-freezing techniques remain hit or miss. Few labs, for example, have been able to routinely repeat the success that Ryuzo Yanagimachi of the University of Hawaii, Honolulu, has had in reconstituting strains from frozen germ cells. TJL staff “have been trying for a year and can't produce a mouse,” says Mobraaten. To overcome the obstacles, NIH is funding a special mouse reproduction initiative that Mobraaten says “is looking promising.”

    But even freezing sperm may have limitations. “It seems to work very well for strains that have a mixed background,” says Mobraaten, but inbred strains don't do well. “If you have a valuable [inbred] strain, I wouldn't rely on it.”

    Ironically, Mobraaten notes, the freeze-storage plans now viewed as a form of salvation once were criticized as extravagant. “The complaint early on was that we were going to create a mouse museum” of unused strains, he says. Today, however, TJL—which stores about 1000 strains as embryos—is “recovering to the tune of 150 strains a year.”

    The mouse redefined

    While mouse experts are confident that they can leap technical hurdles, some worry that future animal rights issues may be more difficult to surmount. Traditionally, mice have slipped “under the animal rights activists radar screen—they just don't have the sympathy factor generated by a dog or chimp,” says one industry executive. But that is changing. In Europe, groups are pushing the Council of Europe to more stringently regulate mouse use. And in the United States, breeders are keeping a close eye on a bid by animal rights groups to have the mouse redefined as a “regulated animal” under the U.S. Animal Welfare Act (AWA) (Science, 5 February 1999, p. 767b), which currently exempts mice, rats, and birds from caging and inspection requirements.

    If the effort is successful—and preliminary signs are that it will be—mouse breeders and researchers may have to submit to new caging rules that could reduce colony densities. TJL's Cook, for one, worries that such rules could increase researchers' costs “by 20% or more.” But Charles River shrugs off the threat that increased AWA regulation could pose to its business, noting that competitors would all have to play by the same price-raising rules. And Taconic's Phelan is philosophical about regulatory changes, viewing them as one of many winds buffeting the mouse economy. “This is a very rapidly changing and maturing business,” says Phelan. “We're doing things now we wouldn't have dreamt possible a few years ago. We just have to get used to the fact that when it comes to mice, we're dealing with a whole new world.”


    A Mouse Chronology

    1. Elizabeth Pennisi

    1664 Robert Hooke observes the reactions of mice in experiments on air, the first recorded use of mice in scientific research.

    1900 Retired schoolteacher Abbie Lathrop begins breeding “fancy” mice at her farm in Granby, Massachusetts. Initially sought as pets, the Granby mice become important in research.

    1908 William Castle opens Harvard's Bussey Institution, where many early mouse geneticists get their start.

    1909 Clarence Little begins to develop the first inbred strain, designated DBA for dilute, brown, and non-agouti.

    1914-19 Lathrop sends mice that developed tumors to Leo Loeb at the University of Pennsylvania, who publishes pioneering papers on cancer.

    1913-16 Halsey Bagg develops the BALB/c (Bagg albino) mouse for behavioral experiments.

    1915 J. B. S. Haldane et al. Publish the first genetic linkage study, establishing the linkage between two coat-color mutations.

    1919 Mouse genetics research begins in earnest at the Cold Spring Harbor Station for Experimental Evolution.

    1921 L. C. Strong breeds a Bagg albino with an albino from Little's stock and starts the first of many tumor-prone strains, called the A strain, known for mammary and lung tumors.

    1921 Using a pair of black mice from the Granby farm, Little develops the C57BL and C57BR strains.

    1928 L. C. Dunn breeds Strain 129, which later proves to have a high incidence of testicular cancer; the strain is now valued as a source of embryonic stem cells for making knockout mice.

    1929 Little starts The Jackson Laboratory in Bar Harbor, Maine, with help from Detroit industrialists who had previously recruited him to the University of Michigan.

    1937 Peter Gorer shows in mouse studies at The Jackson Lab that transplant rejection is primarily governed by what he calls the H2 genetic locus, later described as the major histocompatibility complex, a key component of immunity.

    1939 International Committee on Standardized Nomenclature for Mice begins, bringing order to the naming of mice and their genes.

    1947 Britain launches the Medical Research Council (MRC) Radiobiology Unit—now known as the MRC Mammalian Genetics Unit and the U.K. Mouse Genome Centre—in Harwell, U.K., using radiation to carry out large-scale mutagenesis experiments. Harwell becomes Europe's hotbed of mouse genetics.

    Researchers at Oak Ridge National Laboratory in the United States also do radiation studies. The mutant mice lead to major advances in mouse genetics.

    A fire destroys most of The Jackson Lab and its mice. Researchers rally to rebuild stocks.

    Late 1940s George Snell develops congenic strains of mice—identical but for a small chromosomal segment—by breeding for differences only at the H2 locus. This opens new areas of immunological research and earns Snell a Nobel Prize.

    1949 The informal Mouse News Letter begins its 40 years of publication under that name. At its peak, some 60 labs contribute to it.

    1950 Obese mouse is discovered at The Jackson Lab. The first animal model for obesity, the mouse later proves to have a key mutation in the leptin gene.

    1954 Leroy Stevens develops an ovary transplant procedure that enables mutant strains to be propagated even if the mutation causes the animal to die before it can reproduce.

    1958 Margaret Green at The Jackson Lab starts a card-file database of mouse linkages and loci, which forms the foundation of the Mouse Genome Database. Eventually, the National Institutes of Health (NIH) begins supporting the database.

    1961 Harwell's Mary Lyon proposes X-chromosome inactivation, in which one chromosome in an X-chromosome pair shuts down to maintain the right balance of gene activity.

    1962 The nude mouse, lacking hair, is discovered in Ruchill Hospital, Glasgow, U.K. Several years later, scientists realize that its lack of a thymus means it produces no T cells. It becomes an important tool for immunological studies.

    c. 1970 Richard Gardner of Cambridge, U.K., performs surgery on mouse embryos, opening the way to embryo transfer, embryonic stem cell research, and transgenic mouse technology.

    1971 Donald Bailey develops the first recombinant inbred strains of mice by crossing two inbred strains. The resulting inbreds prove useful for genetic mapping and gene hunting.

    1972 U.K. researcher David Whittingham shows that frozen mouse embryos can survive thawing, making it possible to preserve strains without continuous breeding.

    1976 Rudolf Jaenisch, now at the Massachusetts Institute of Technology (MIT), uses a virus to transfer DNA to mouse embryos, the first report of success in creating a transgenic mouse.

    1978 François Bonhomme in France breeds two species, Mus spretus and Mus musculus, enabling geneticists to build the first comprehensive linkage map of the mouse genome. This makes the mouse a “formidably efficient system for genome mapping,” notes mouse geneticist Phil Avner.

    1979 William Russell of Oak Ridge proves that the chemical ethylnitrosourea (ENU) is effective in generating mouse mutations. Oak Ridge and other labs that had been studying radiation effects begin producing ENU mutants.

    1979-80 Using microinjection to insert DNA into a mouse egg, six labs independently demonstrate that foreign DNA can be put into the mouse genome.

    1981 Martin Evans and Matt Kaufman in Cambridge, U.K., isolate mouse embryonic stem cells, which can develop into the full range of tissues.

    1982 By inserting rat growth hormone gene into a mouse, R. D. Palmiter et al. create an extra-large transgenic mouse—and a media splash. The same year, U.S. officials loosen restrictions on DNA cloning in mammals, and the book Molecular Cloning: A Laboratory Manual ushers in the era of transgenics.

    1983 The SCID mouse, which lacks an immune system, is discovered and becomes a valuable tool for studying human tumors transplanted into mice.

    1984 Joseph Nadeau and Ben Taylor's analysis of 83 genes in mice and humans indicates that the mouse genome is an extremely good model for the human genome—but with 150 rearrangements.

    1985 Brian Sauer's introduction of the Cre-loxP system for temporal control of transgenic gene expression draws little attention at San Francisco meeting, but 5 years later causes quite a stir when he and DuPont obtain a patent on it.

    1985 Harwell's Bruce Cattanach describes genetic imprinting in mice, an epigenetic phenomenon now known to occur in humans as well. Imprinted genes are differentially expressed in the offspring depending on the parental origin of the chromosome.

    1987 Mario Capecchi's team at the University of Utah describes a method for making knockout mice, as does Oliver Smithies's group at the University of Wisconsin.

    To Make a Knockout Mouse

    Introduce a designer gene into mouse embryonic stem (ES) cells in culture.

    Screen ES cells and select those whose DNAincludes the new gene.

    Implant selected EScells into normal mouse embryos, making “chimeras” of mixed heritage.

    Implant chimeric embryos in pseudopregnant females.

    Females give birth to chimeric offspring, which are bred to verify transmission of the new gene, producing a mutant mouse line.

    1988 Harvard mouse patented.

    1990Mouse News Letter becomes a peer-reviewed journal, Mouse Genome, marking an increase in formality in the mouse community. In 1997, that journal is folded into Mammalian Genome.

    1992 Researchers at MIT and at Baylor College of Medicine in Houston describe a knockout mouse lacking the p53 tumor-suppressor gene, an instant sensation among researchers.

    1992 The U.S. District Court rules that mice, rats, and birds are not excluded from the Animal Welfare Act of 1971. Although the ruling has no immediate impact, activists are now arguing that the decision requires stricter controls on rodent use.

    1993 The NIH starts supporting a new repository to make genetically engineered mutant animals widely available to the research community. With molecular geneticist Harold Varmus at the helm, NIH takes even more notice of mice. In 1998, Varmus stimulates a Trans-NIH Mouse Initiative.

    1996 Eric Lander's group at MITpublishes a map of the mouse genome with more than 7000 markers.

    1997 Merck Genome Research Institute funds the creation of 150 new mutant mouse types at Lexicon Genetics for restriction-free distribution to the basic research community.

    1998 Ryuzo Yanagimachi's team in Hawaii clones mice from somatic cells by using nuclear transfer and discovers how to freeze-dry sperm for future use.

    1998 Researchers in Munich, the United Kingdom, and, later, Australia, launch large-scale ENU mutagenesis projects to provide the research community with thousands of new mutants by 2001.

    1999 In Japan, Yoshihide Hayashizaki's group determines the first set of full-length mouse complementary DNAs, 20,000 of which have been put on microarrays for analyses of gene expression. NIH eventually gains access to the full database for intramural scientists; others hope to do the same.

    2000 Mouse genomics takes off.

    Sequencing the Mouse Genome

    The United States has awarded $130 million through 2001 to begin sequencing the mouse genome, and 10 U.S. centers have taken on the task of developing maps, generating some whole-genome shotgun sequence data, and sequencing biologically important pieces of DNA. The U.K.'s MRC is providing funds for the sequencing of 50 million bases of the mouse genome. In February, GenBank had about 1.2% of this 3-billion-base genome in-house, more than half of that as a rough draft. The goal is to have a rough draft by 2003 and a finished genome by 2005. (For an update on the sequence, see In April, Celera Genomics began sequencing the mouse on its own.


    Decoding a Mouse Name

    1. David Malakoff

    129S7/SvEvBrd-Hprtb-m2 For mice used in science, pedigree can be everything. So researchers have developed a precise naming code that tells users a bit about each mouse's history, source, and traits. Major breeders, for instance, begin names with their own acronym—Cr for Charles River, HSD for Harlan Sprague Dawley, Tac for Taconic, and J for The Jackson Laboratory—and then add strain information. Shown above is the name for one strain of the popular 129 mouse. It tells users that this is the #7 substrain with a steel-colored coat (S7), and that it has passed through labs run by researchers named Stevens (Sv), Evans (Ev), and Bradley (Brd). Finally, the name denotes a mutation on the “b” allele of the Hprt gene, with “m2” showing that it was the second mutation of that allele.

    Names are getting longer as researchers demand more information on genotype and phenotype. “We live and die by the names,” says Taconic's Sam Phelan, but “they are having a hard time keeping up.” Soon, he says, names will be just the tip of an information iceberg, as researchers routinely turn to large electronic databases to get the complete skinny on their mouse model.


    The Mouse House as a Recruiting Tool

    1. Gretchen Vogel

    Talent hunters at major research centers are luring scientists by promising to build state-of-the-art animal facilities and reduce cage charges

    Although several universities have tried to recruit developmental neuroscientist Susan Ackerman, she has rebuffed them all. They've offered her generous salaries and state-of-the-art labs, but they can't match the most important perk: the unusually low cost of caring for mice at her current institution, The Jackson Laboratory (commonly known as “Jax”) in Bar Harbor, Maine. The cost of mouse care at one university, she says, “was going to be far more than my salary.” This would have limited her ability to create the genetically altered animals she uses to study how the nervous system is wired during development. Having more animals means you can test more ideas, and Ackerman says, “being at Jax allows me to do more risky experiments.”

    Ackerman is not alone in sizing up jobs according to the mouse factor. Mouse geneticist John Mercer says he made his first job decision almost solely on mouse costs. The two offers he was considering were similar, he says, except for charges at the animal-care facility. The University of Texas Southwestern Medical Center (UT Southwestern) in Dallas charged researchers 48 cents per day per cage (a cage holds up to five mice), whereas the other university charged 26 cents per day per mouse. That made his decision simple: He accepted the job at UT Southwestern.

    Within a year, however, Mercer's careful analysis went out the window as UT Southwestern's costs doubled, and he began comparing facilities again. In 1995, Mercer moved to his current job at the McLaughlin Research Institute in Great Falls, Montana, where he pays as little as 18 cents per cage per day. “It's like getting a grant that can never be taken away,” he says. The bargain rates have allowed him to try more frequent and more daring experiments, and at McLaughlin he's created several useful knockout mice.

    For many scientists, the subject of animal costs may never come up, but for geneticists, developmental biologists, immunologists, neuroscientists, and others who use mice as models, it is a major concern. Indeed, a recent committee at the National Academy of Sciences listed inadequate funding for mouse care as one of the top threats to immunology research in the United States.

    Developmental biologist Brigid Hogan of Vanderbilt University in Nashville, Tennessee, says she uses the bulk of her Howard Hughes Medical Institute funding to pay for animal care. For her, a generous animal budget is essential. To help colleagues track the issue, she set up a Web site that compares mouse-care costs at several institutions,* as reported by researchers. But some have done more than report on their troubles.

    At several universities, frustrated scientists whose mouse-care bills have skyrocketed have banded together to demand that administrators give an explanation. Some found that they were subsidizing research on larger, more expensive animals, says immunologist Irving Weissman of Stanford University. Several years ago, he and several colleagues asked Stanford to account for the actual costs of keeping each type of animal. Once the results were in, he says, the university lowered mouse charges more than a third and raised charges for other animals.

    “Before the rate change at Stanford, I had to raise $800,000 to $1 million a year to keep the 2000 to 3000 cages I believe I need for the research I do,” Weissman says. “That meant I was spending most of my time writing grants.” Other researchers, he says, had to decide between giving up mouse research or leaving Stanford.

    A combination of factors drove animal costs dramatically higher over the last few years, says Linda Cork, head of Stanford's Department of Comparative Medicine, which oversees animal care. The main problem was the federal government's decision to classify animal-care buildings as “specialized facilities,” as they were used by only a subset of researchers. This meant that universities could no longer pay for their construction or maintenance with the “indirect cost” allowance that pays for labs, libraries, and infrastructure. Institutions compensated in various ways. Some found the funds in departmental budgets, Cork says, but many others decided to pass costs along to the researchers who used the buildings.

    At the same time, managed care began to squeeze medical school budgets, drying up funds—including money for animal care—that had helped underwrite research. All the while, scientists were producing new and intriguing animal models, driving up the demand for transgenic mice. The result: Animal-care costs rose across the board.

    But there is some relief in sight. The National Institutes of Health decided last year to return to an earlier policy and allow universities to include animal research facilities in the indirect cost rate. Cork believes the change will enable many institutions to significantly lower the daily charges for keeping mice. It will take time to reach some researchers, however, because universities renegotiate their indirect cost rate only every 5 years.

    Universities are also responding on their own. Nearly 40% of those in a recent Yale survey said they were planning new animal facilities. Baylor College of Medicine in Houston, Texas, for example, is in the final stages of constructing a building designed to house 45,000 mouse cages. The project includes several cost-cutting innovations, says Bob Faith, director of Baylor's Center for Comparative Medicine. For example, Baylor hopes to save on labor costs by using conveyor belts and robots to clean cages. And each cage will have a constant stream of fresh air, which will not only help prevent disease but also reduce the need for fresh bedding. When the new facility is completed, he says, the university will actually lower its daily cage rates, from 31 cents to 26 cents per cage.

    It's a step in the right direction, says Weissman, but he thinks more universities need to follow suit. “As long as artificially high prices for mouse care exist,” he says, this obstacle, “not the right-to-life or animal-rights [movements], will be the major stumbling block for the transfer of molecular biology to humans.”


    A Deluge of Patents Creates Legal Hassles for Research

    1. Eliot Marshall

    Scores of animals have been patented since Harvard claimed the OncoMouse in 1988, but now Merck and NIH are funding patent-free mice

    Tom Doetschman, a geneticist who creates exotic strains of mice, says he's beginning to feel “old-fashioned.” It's not that his methods are antique; far from it. The animals he breeds for genetic research are in high demand, and his lab at the University of Cincinnati (UCI) has a hard time keeping up with requests. Doetschman has created over 120 knockout (gene-deleted) mice in the past decade, he says, and given them away at cost. Unlike peers who have patented mice with ailments that mimic everything from AIDS to bovine spongiform encephalopathy or “mad cow disease,” he has never patented an animal. “I make the mice available to anyone who wants them—no questions asked, no restrictions, nothing,” he says. It is this noncommercial attitude that makes Doetschman feel that he's in “an incredible minority.”

    To Doetschman, the mice are tools to be shared. But to UCI's technology transfer chief, Norman Pollack, they are university property. Pollack understands Doetschman's view: “In practice I don't have a problem with it,” he says, partly because engineered mice are not great moneymakers. But in principle, Pollack cannot agree that a faculty member “has the right to give that stuff away.” Recently, UCI warned Doetschman that he may be giving away mouse technology patented by others.

    This tension between the creators and the controllers of knockout mice is indicative of a tension throughout the research world. Pollack is one of thousands of university officials empowered under federal law—the Bayh-Dole Act of 1980—to capitalize on federally funded research. Many have leapt at the chance, even if it has meant selling inventions to other researchers. And a new generation of scientists assumes that research tools will be marketed.

    But commercialization has brought with it legal problems, including high attorneys' fees. For example, Elan Pharmaceuticals of Dublin, Ireland, is now locked in a bitter fight in U.S. federal court in San Francisco with the Mayo Foundation over rights to a mouse with Alzheimer's symptoms. The tussle has roiled the aging research community for more than a year. And in other fields, scientists seeking custom-engineered mice have complained loudly about the tough licensing conditions and high prices of animals offered by Lexicon Genetics Inc. of The Woodlands, Texas. Many scientists, as users of these tools, worry that the tendency to patent every new increment of genetic discovery, including every new mouse, if not resisted, could impede genetic medicine. This has led to a backlash aimed at freeing research tools, especially mice, from commercial red tape. The effort began with individual scientists, was taken up by the National Institutes of Health (NIH), and has been joined by at least one major pharmaceutical company.

    Privatizing mammals

    Harvard University began the scramble for genetic mouse property in 1988. That's when it obtained the first transgenic animal patent, U.S. patent number 4,736,866, for a “non-human eukaryotic animal whose germ cells and somatic cells contain an activated oncogene sequence introduced into the animal, or an ancestor of the animal, at an embryonic stage.” Broadly interpreted, the invention by Philip Leder of Harvard and Timothy Stewart of Genentech Inc. in South San Francisco covers any animal genetically engineered to produce tumors. Harvard gave DuPont an exclusive license to distribute the tumor-prone mice but retained the right to use them freely in its own research.

    The Harvard mouse fired up a smoldering debate on whether it is right to patent life. The Supreme Court had already ruled in 1980 (Diamond v. Chakrabarty) that General Electric could patent an oil-digesting bacterium because it had been genetically engineered and was not a product of nature. Church groups and animal rights organizations argued that this policy, if extended, would lead to a devaluation of life. The debate simmered on, and for 4 years, the U.S. Patent and Trademark Office had an unofficial moratorium on animal patents. Then it plunged ahead in 1992, awarding three patents on mice and one on a disease-resistant chicken in a single year. The pace picked up in the 1990s, hitting a peak at 47 patents issued in 1997.

    But other patent offices were slow to follow. The European Patent Office (EPO), for example, only received permission in principle to patent animals in 1998, after a 10-year public debate. And in December 1999, an EPO appeals board officially affirmed that patents on plant varieties are permitted—“a beautiful decision,” according to assistant U.S. patent commissioner Stephen Kunin. Today, he says, “Europe is operating along U.S. lines,” as is Japan. The clear exception is Canada. Its patent office rejected the OncoMouse patent in 1993, and Harvard has been battling ever since to reverse the decision. Harvard, unsuccessful so far, is taking the case to Canada's supreme court this year.

    The mouse patenting frenzy didn't upset basic researchers initially. After all, it was they who started it. But many became outraged by the consequences of patenting—particularly by the prices and proprietary restrictions on the use of mice.

    One angry response came from a Nobel Prize-winning scientist in oncogene research at the University of California, San Francisco (UCSF): Harold Varmus. He helped organize the mouse malcontents in 1992 and 1993. As Varmus recalls, he and Douglas Hanahan, another UCSF scientist, thought the prices and conditions on use of the p53 knockout mouse—then supplied by a company called GenPharm, which was acquired by Medarex Inc. in October 1997—were “abhorrent.” GenPharm was charging $80 to $150 per mouse and forbidding academics to breed the animals. So, Varmus says, “we went on the warpath.” Varmus held an impromptu meeting at the Cold Spring Harbor Laboratory mouse genetics meeting in 1992. About 300 aggrieved scientists showed up and began talking revolution (Science, 2 April 1993, p. 23).

    This gathering led to a review of restrictions on the sharing of research tools at the U.S. National Academy of Sciences in Washington, D.C., in March 1993. The NIH followed up in 1993, just before Varmus was appointed director, with funding for a new shared mouse facility. Together with private donors, NIH backed the Induced Mutant Resource at The Jackson Laboratory (widely known as “Jax”) in Bar Harbor, Maine, a repository of genetically altered mouse strains that was meant to give all researchers equal access to new genetic research tools (see main text and

    The repository helped. But there were logistical problems—and new legal barriers. Jax couldn't afford to maintain live stocks of all the animals researchers wanted to share; space and resource constraints made it necessary to keep many strains as frozen embryos. The lab began having big headaches over the fine print in conditions that were placed on who could or could not receive animals from its repository.

    In the mid-1990s, Jax stopped handling mice created with a popular gene-insertion method known as Cre-loxP, which allows the experimenter to set conditions that cause a gene to be turned on or off. In 1990, DuPont had obtained a patent on mice incorporating this method and made itself unpopular by demanding that researchers not share the technology among themselves without the company's prior approval. DuPont also contacted scientists who had published data from Cre-loxP animals and asked them to sign an agreement stipulating that DuPont could review their scientific articles before publication. Furthermore, the company sought “reach-through” rights, or rights to second-generation inventions that might arise from using these animals. “It was a major problem,” says David Einhorn, Jax's legal counsel: “Nobody was able to exchange materials” freely any longer.

    Varmus again intervened, this time from a position of greater influence. As NIH director, he refused in 1997 to sign an agreement with DuPont on the Cre-loxP mouse on behalf of NIH, making it impossible for thousands of intramural staffers at the NIH campus in Bethesda, Maryland, to get access to the technology. It was a nuisance for them and an embarrassment for DuPont, but it produced a change. Varmus wrote to DuPont that the company's restrictive terms could “seriously impede further basic research and thwart the development of future technologies that will benefit the public.” After a year of negotiation, DuPont made concessions: The company did away with demands for prepublication review for research-only uses of Cre-loxP mice, loosened up animal sharing provisions, and dropped the reach-through property claims for NIH-based scientists (Science, 28 August 1998, p. 1261).

    In December 1999, DuPont reached another agreement with NIH on mouse rights—again through the intervention of Varmus. After hearing a plea from Varmus that it relax its rules for use of the OncoMouse, DuPont said that NIH scientists and NIH grantees at nonprofit institutions could exchange animals without directly involving the company (Science, 28 January, p. 567).

    Other initiatives now in the works could soon make it easier for all researchers to get access to patent-free transgenic mice. The pharmaceutical firm Merck & Co. Inc. of Whitehouse Station, New Jersey, announced plans last year to spend $8 million to have Lexicon Genetics create 150 patent-free transgenic mice to be made available at cost through The Jackson Laboratory. Fourteen of these model transgenics have been created, 61 more are in the pipeline, and the rest will be designated for production soon by a panel of outside experts, says Thomas Caskey, the recently retired chief of Merck's Genome Research Institute who conceived the project. Caskey says the mice will be shared without patent or use restrictions. He explains that Merck wants to give scientists new tools that have no legal hassles attached. But Merck is not motivated entirely by altruism: Minimizing such property claims will benefit the company as well.

    In a related effort, NIH has committed itself to a multistage “mouse initiative” that will pay to sequence the mouse genome, develop thousands of new model transgenic animals, and characterize the animals' phenotypes. As a policy matter, NIH leaders will insist that people who accept grants to do this work not file patents. NIH rarely takes this step, says Maria Freire, director of NIH's Office of Technology Transfer, but in this case it will invoke an “exceptional circumstances” clause of the Bayh-Dole Act that allows the government to insist that the animals it produces will be patent-free.

    If these new projects pay off, researchers will have access to thousands of new mouse models that have no intellectual property strings attached. And Doetschman may discover that, rather than being old-fashioned, he was ahead of the times.