# News this Week

Science  15 Dec 2000:
Vol. 290, Issue 5499, pp. 829
1. HUMAN GENOME

# Storm Erupts Over Terms for Publishing Celera's Sequence

1. Eliot Marshall

If Celera Genomics—the biotech firm in Rockville, Maryland, that has sequenced the human genome—had decided never to publish its work, nobody would have kicked up a fuss, says company president J. Craig Venter. But last week, when Venter and his team submitted a paper on the human genome to Science, a major row ensued.

The focus of the dispute, which has been raging behind the scenes for weeks, is the conditions under which Celera is prepared to make its data publicly available. The argument went public on 6 December, when geneticist Michael Ashburner of Cambridge University e-mailed an open letter to Science's board of reviewing editors and members of the press. He slammed an agreement on data release that Science had reached with Celera as a condition for accepting its paper for review. Ashburner's letter prompted Science to issue a statement* spelling out the terms of the agreement: Celera would make the entire sequence available free of charge through its own Web site, but place conditions on some uses of the data. Ashburner said he was “outraged and angry” that Science was not forcing Celera to deposit its sequence data in GenBank, a public repository whose data anyone can access and use without restrictions.

This spat is the latest round in an intense rivalry between Venter and leaders of the Human Genome Project, a publicly funded consortium that has produced its own draft human genome sequence. Some researchers involved in the public effort have been among the most vehement critics of Celera's data release policies; Ashburner's open letter expressed concerns many of them have voiced in private. The letter also came at a critical time: The day after Ashburner went public, leaders of the public project met to decide where to submit papers on their work. According to one geneticist, who asked to remain anonymous, they sent them to Nature.

Science Editor-in-Chief Donald Kennedy says the agreement with Celera is the product of months of negotiations conducted in part by intermediaries in the scientific community. “This was a tough call,” he says. The easy way out would have been to set a standard for publication that Celera could not meet, Kennedy says, but that would have been “the moral equivalent” of rejecting the submission arbitrarily. He argues that Celera has a right to protect its investment in generating the sequence; the new data release agreement takes this reality into account while allowing researchers to use a wealth of data that might otherwise be kept private.

The statement Science released says academic users may access the entire sequence on Celera's site, “do searches, download segments up to one megabase, publish their results, and seek intellectual property protection.” For downloads of more than 1 megabase, scientists must submit an agreement, signed by an institutional representative, not to redistribute the data. Commercial users have many of the same rights but must sign a Material Transfer Agreement promising not to commercialize their results or redistribute the sequence data. Alternatively, companies may subscribe for a fee or seek a license.

Kennedy noted that Science has a longstanding policy of asking authors to deposit DNA sequence in a publicly accessible database, and he argues that the agreement with Celera fully upholds that principle. The journal, he says, “has never stipulated a particular database,” although by tradition, it has sent authors to GenBank. Nature's policies don't appear to be much different. Asked whether he would consider a paper from a company that would release data on its own Web site, Nature Editor Philip Campbell said he'd prefer “not to rely on databases at all.” But he acknowledged that “there will be circumstances where it's unavoidable, and I can imagine circumstances where the only database available might be privately owned. … If the community judged the database to be well curated and supported for the long term, and if access conditions were appropriate—as stated in our current policy—we'd accept [the use of a private database].”

Kennedy also noted that the plan includes a “unique departure from our normal procedures”: Science will hold a copy of Celera's data in escrow to reassure the community that the information will remain accessible if Celera should later change its policies.

These assurances don't appease Ashburner, who wrote in his open letter that allowing Celera to avoid depositing its sequence in GenBank will “fragment” genomic data “across many sites.” He predicted that “today's ease of searching [genomic data] will be gone, and gone forever.” He said he has “nothing whatsoever against the idea that Celera sequence the human genome and sell it,” but he said the company also wants “the academic kudos that goes with it.” A former member of Science's board of reviewing editors, Ashburner urged current members to quit and refuse to review Celera's paper.

Other critics of the agreement have been more restrained in their public statements. Harold Varmus, former director of the National Institutes of Health and now president of Memorial Sloan-Kettering Cancer Center in New York City, confirms in an e-mail that he was one of about 15 scientists who wrote to Kennedy in November to express concerns about the Science-Celera discussions. “I remain concerned about the new precedent that may have been set,” Varmus now says. Other people may demand exceptional treatment, he warns, and asks: “What will Science magazine do next time?”

Several biomedical leaders who have been involved in backstage negotiations over the agreement—including Thomas Cech, president of the Howard Hughes Medical Institute (HHMI) in Chevy Chase, Maryland; David Baltimore, president of the California Institute of Technology in Pasadena, California; and Bruce Alberts, president of the National Academy of Sciences (NAS)—now see merits in the agreement, at least for academic users.

Baltimore says he has reviewed the terms as they apply to academic institutions and, “in my amateur opinion,” they are acceptable. He adds that Kennedy has “done a great service to craft an agreement that allows the door to be opened” to privately held data. Cech, responding to questions by e-mail, said the terms for academics are “very close to being acceptable,” adding that some “vague passages” need to be clarified before HHMI investigators would be permitted to sign up to see the data. But “we do not expect these to be controversial.” At NAS, Alberts says he's been assured by scientists whose judgment he trusts that the data-sharing provisions for academic researchers are satisfactory. He adds that it makes sense to try to work with private companies, partly because they will be doing a “massive amount” of DNA sequencing in the future.

Terms for commercial users of Celera's data are another story, says Cech, who argues that they are so restrictive they might “exclude users in the for-profit arena.” Similar concerns were raised by one leader in the public consortium, who asked to remain anonymous.

At some point, the arguments over access to Celera's data could become moot. Alberts suggested in a public statement that the sequence Science will hold in escrow should be turned over to GenBank “once a sufficient amount of time has elapsed to allow Celera to protect its legitimate business interests.” Venter says he's willing to consider that option. In the next couple of years, says Venter, “we will definitely revisit that suggestion and see if it makes sense.”

2. U.S.-RUSSIA TIES

# Spy Conviction Strains Science Collaborations

1. Richard Stone

CAMBRIDGE, U.K.— The conviction last week in Russia of U.S. businessman Edmond Pope on charges of espionage may add to already growing tensions in scientific collaborations between the two countries, according to officials on both sides. The recent strains appear to be a reaction to a broad range of national security concerns in each nation.

In Russia, pressure is coming from the increasingly assertive Federal Security Service (FSB), the successor to the Soviet KGB. In the United States, security breaches at the national laboratories and throughout the intelligence community have led to restrictions on visiting scientists from a handful of countries, including Russia, that are deemed “sensitive.”

The heightened concerns have put a crimp in U.S. efforts to reduce Russia's proliferation threat by linking U.S. scientists with Russians at dozens of once top-secret defense research centers. These efforts include programs such as the Department of Energy's Initiatives for Proliferation Prevention and the multinational International Science and Technology Center. “Many of the programs that [Defense Department researchers] are involved in are stopped. Many visits to Russia are postponed indefinitely,” says one U.S. government official who spoke on condition of anonymity. Added another official, “We are concerned about the situation and its dampening effect on scientific cooperation.”

Scientific exchanges have also been affected. The State Department's Bureau of Consular Affairs, for example, imposed a 2-month clearance of all Russian participants—twice as long as it took last year—for an October workshop on dangerous pathogens held at the Sandia National Laboratories' Cooperative Monitoring Center in Albuquerque. The new policy “has led to the cancellation of many foreign visitors,” says one official. “We may be seeing some tit for tat,” adds a second official.

U.S. scientists visiting Russia, meanwhile, face more delays in entering institutes or areas closed to the general public. They are also experiencing more incidents in which the FSB or border guards have confiscated equipment deemed sensitive, such as Global Positioning System receivers. Nonprofits that work with defense scientists have also noticed a chillier climate. “The rules have changed,” says Gerson Sher, director of the Arlington-based Civilian Research and Development Foundation. “We're seeing a trend toward more rigor” in how applied, market-oriented projects are administered in Russia.

The 20-year sentence meted out to Pope, imprisoned for 8 months after being accused of buying secret information on a high-speed torpedo that Western experts say has been sold openly to other countries, has added to the strains. Pope, who is in poor health, was expected to receive a presidential pardon and be released from prison soon after Science went to press. However, his treatment has riled U.S. officials, who have asserted Pope's innocence from the start.

The Pope case highlights the FSB's resurgence under Russian President Vladimir Putin. The security service suffered a pair of blows in the last year, when Russian ecologist Vladimir Soyfer and former Navy officer Alexander Nikitin, both accused of revealing classified data, won key court victories (Science, 10 March, p. 1729). The Pope verdict “suggests a revitalized FSB and a danger to all researchers,” says Paul Josephson, a Russian historian at Colby College in Waterville, Maine.

Pope's release would ease tensions—but only a bit. U.S. officials are on tenterhooks after Russia threatened last month to resume arms sales to Iran after a 5-year hiatus on new contracts. “The worst-case scenario,” says one observer, “is that all the technology cooperation programs are halted.” Others have a more hopeful attitude. “I continue to be an optimist,” says Lev Sandakhchiev, director-general of the State Research Center of Virology and Biotechnology “VECTOR” in Novosibirsk, Russia, a former bioweapons lab. “Russia and the U.S.A. are bound to have good relations.”

3. ECOLOGY/PALEONTOLOGY

# Colorado River Clams Provide Benchmark

1. Mark Muro*
1. Mark Muro writes from Tucson, Arizona.

When naturalist Aldo Leopold explored the Colorado River delta in 1922, he found a “milk-and-honey wilderness.” But 27 years later, he wrote that “I am told the green lagoons now raise cantaloupes.” Conservationists have long contended, largely in impressionistic terms, that 70 years of American dam building and water diversion have destroyed the biological richness of the delta, a key nursery of marine life at the end of the Southwest's great watercourse. Now researchers have confirmed those suspicions, using an important ecological player as a quantitative marker.

“Basically, we've used clam shells to quantify what things were like before the dams and found they were vastly different,” says Karl Flessa, a geoscientist at the University of Arizona in Tucson who led the four-university team's work, which is reported in the December issue of Geology. The work, says Sally Walker, an invertebrate paleontologist at the University of Georgia, Athens, “shows paleontology can be extremely useful for solving environmental questions by establishing an ecosystem's long-term past before humans altered it. That's powerful.”

Flessa and his colleagues in Virginia and Mexico studied clams in the delta because, unlike other animals that decay and are lost to the geological record, clams leave behind hard shells to tell of past abundance. Clams furthermore stand as a “proxy” for “the whole marine ecosystem and its health,” says Flessa, who notes that numerous fish, mammals, and migratory shorebirds depend directly on them for food. Flessa and his colleagues hoped that an analysis of the vast islands of gleaming white shells, using paleontological, geochemical, and geochronological methods, would allow them to estimate the delta's biological productivity both before and after the river's water was diverted.

To do so, the researchers carried out a series of simple mathematical calculations. First, they used satellite images, trenches excavated in shell-rich beaches, and field measurements of ridge density to estimate that the remains of some 2 trillion clams lay entombed in great shell ridges and islands in the delta. Then they dated 125 shells by analyzing changes in their amino acids and calibrating the results with radiocarbon dating. Virtually all those 2 trillion shells accumulated over the 1000 years from A.D. 950 to 1950, they found. Finally, they used stable isotope profiles recorded in shells to calculate the population turnover rate, which allowed them to calculate that 6 billion mollusk bivalves flourished at any given time in the area. From that number, they calculated an average density of 50 clams per square meter over the last millennium. In contrast, earlier this year seven sample areas yielded estimates of just three individuals per square meter.

Michal Kowalewski, a geobiologist at Virginia Polytechnic Institute and State University in Blacksburg and one of the project's leaders, believes that the productivity of the delta system has fallen at least 95% since the 1930s, when Hoover Dam was built. “That's a big drop, but in fact our calculations are so conservative it's probably much worse than that—maybe 10 times worse,” Kowalewski says. He blames reduced fresh water and nutrient flows to the delta. About 90% of the river, or about 13.5 million acre-feet of water a year, is now diverted to support the fields and booming cities of the Sun Belt.

The new work could have both local and global implications. In the Southwest, the clam counts could help environmentalists secure greater water flows to the delta to restore its species. “You need numbers to negotiate with and litigate with, and [Flessa's work] gives us numbers,” says environmentalist David Hogan of the Tucson-based Southwest Center for Biological Diversity, which has been active on the issue.

The clam research also offers a dramatic example of how the methods of paleontology can be used to address environmental problems elsewhere, suggests Walker. Techniques like Flessa's and Kowalewski's can provide quantitative historical baselines even when long-term ecological monitoring cannot, she says. “Applying methods like these can give you a numerical sense of the scale of what has happened and then a metric, or benchmark, for attempting remediation,” says Walker.

To Flessa, the numbers provide mute testimony on “what has been lost” during 70 years of aggressive water management in the region. He says that the federal dam builders in the Southwest too often ignored the costs of irrigating fields and slaking the thirst of sprawling desert cities. “Now,” he says, “we're providing some quantitative assessments of those impacts. That they're huge will help, I hope, to sharpen future policy.”

4. PLANT BIOTECHNOLOGY

# Italian Scientists Blast GMO Restrictions

1. Lone Frank*
1. Lone Frank is a freelance writer in Copenhagen, Denmark.

COPENHAGEN—While plant scientists around the world celebrate the complete sequence of the genome of the mustardlike plant Arabidopsis thaliana (see p. 2054), embattled colleagues in Italy are protesting new rules that bar all field trials involving genetically modified organisms (GMOs). The researchers hope to turn the prevailing tide by bringing their plight to the attention of colleagues around the world and exerting pressure on their government through a petition drive. “It makes no sense to do research related to agriculture if field tests are forbidden,” says molecular biologist Angelo Spena of the University of Verona.

Biotech critics have had a field day in Europe, where resistance to transgenic crops has influenced policy and crimped research funding (Science, 4 February, p. 790). But “only in Italy [are individual scientists] being penalized as a consequence of public concerns,” says biologist Roberto Defez of the National Research Council in Naples. Plant researchers aren't the only ones crying foul. “The issue reaches far beyond biotechnology,” claims physicist Giorgio Benedek of the University of Milan-Bicocca, who cites “a general concern in Italy about this antiscience attitude within the government.”

At the center of the controversy is Agriculture Minister Alfonso Pecoraro Scanio, a Green Party member who took office last April. A longtime critic of transgenic crops, Pecoraro Scanio claims that GMOs pose a threat to human health and the environment. His first strike at research came in July, when he informed the ministry's chief research coordinator, Francesco Salamini, that funding for projects at 23 institutes under the ministry—which carry out the bulk of the country's ag-biotech research—would only flow after a written declaration from researchers pledging that they would not conduct field trials of GMOs.

The next blow came in September, when Pecoraro Scanio issued a new requirement for long-term projects approved since 1996, many involving ongoing or planned field trials of GMOs. According to Defez, the minister “asked individual scientists to modify their original research proposal to remove every aspect concerning use of GMOs.” Only those who complied had their funding renewed. One victim, the first-ever field trial on grapes engineered to taste better, has been halted in Sicily. Such a policy appears to conflict with European Union law, which permits field trials of genetically modified crops that meet restrictions such as adequate safeguards against the spread of transgenes to wild relatives and unaltered crops.

The ministry has also put the kibosh on new research involving GMOs, having declined to approve any proposals since July. According to Defez, a commission composed of representatives from several ministries, including Agriculture, that is responsible for approving field trials “simply postpones applications until it's too late for planting.” Many plant biotech lab studies are in vain if not followed up with fieldwork, claims Spena, who says it would be ridiculous to spend years and considerable funds on creating transgenic plant varieties, only to abandon them because of a flawed policy.

Defez and two colleagues have drafted a petition highlighting their plight. Published on 5 November in the financial journal Il Sole 24 Ore, the petition has garnered more than 1000 signatures so far, including all major Italian scientific societies and notables such as Nobel Prize winner Renato Dulbecco of the Institute for Biomedical Technologies in Milan. Late last month, the American Phytopathological Society became the first international society to sign on.

The Agriculture Ministry insists that scientists are blowing the situation out of proportion. “GMO research is supported in Italy, except in open field trials,” says ministry spokesperson Triantafillos Loukarelis. Scientists, however, assert that Pecoraro Scanio is using the Greens' political leverage to force other government ministers to back his anti-GMO policies. “As a Green fundamentalist, he is blackmailing the rest of the government who depend on the Green vote,” contends Spena, who says any politicians who cross Pecoraro Scanio risk bringing down the government if the Greens were to pull out. “An open society cannot allow science to become subject to the whims of individual ministers,” Spena says. The fight could continue until the next Italian election, expected in summer 2001. “If the minister retains his position,” predicts Defez, “we would see a regular exodus of scientists in biotechnology to other countries or other fields of research.”

5. CHEMICAL PHYSICS

# Magnetic Wires Promise Giant Step for Memory

1. Robert F. Service

The race to cram ever more data onto computer hard drives could soon be veering onto more interesting terrain. So far the track has been level: Hard drives store bits of data in tiny beams of magnetic material lying side by side on a flat disk. By continually shrinking those beams and improving the devices that read and record magnetic traces in them, computer engineers have boosted storage capacity nearly 100-fold over the past decade. In theory, they could pack in even more bits by standing the beams upright. But the leading technology for making dense arrays of magnetic posts requires chemicals corrosive to other disk materials and is limited to making posts of just one size. Now, a better way to make such an array may be at hand.

On page 2126, researchers at the University of Massachusetts, Amherst, IBM's T. J. Watson Research Center in Yorktown Heights, New York, and the Los Alamos National Laboratory in New Mexico describe a potentially cheap and simple method of creating porous plastic templates. By filling the pores with magnetic materials, they can make magnetic posts so small and close together that 1012 of them fit in a square centimeter. If each post could be addressed individually—a feat beyond the capability of today's read-write heads and likely a major challenge—disk drives would be able to store a terabit of data per square centimeter, a 300-fold improvement over current models.

“I think it's going to be important,” says Ivan Schuller, a physicist at the University of California, San Diego, who studies the properties of ultrasmall magnetic structures. And magnetic storage may be just part of the story, says Tom Russell, who led the University of Massachusetts portion of the team. By tweaking the recipe for the template's chemical precursors, he points out, one can create films with pores of different sizes. That may open the door for nanowire templates and arrays that serve as novel porous membranes for chemical separations or as tiny crucibles for controlling chemical reactions.

Currently, laboratories make the most densely packed arrays of tiny, parallel magnetic wires by chemically pitting a thin block of aluminum with tightly packed holes and then filling the holes with a magnetic material such as cobalt or iron. The trouble, Russell says, is that creating the holes requires caustic reactions that can wreak havoc on other components in computer circuitry.

To overcome that problem, Russell and his colleagues turned to two-part plastic molecules called copolymers. A copolymer molecule contains two portions sewn together in the middle, like a strand of spaghetti that is egg-flavored at one end and spinach-flavored at the other. When large numbers of copolymers are mixed together in solution, similar ends quickly crowd together. Researchers have long used this self-segregating property to make thin plastic films in which the copolymer halves arrange themselves into various patterns, including arrays of cylinders of one polymer component standing upright within the other. Those cylinders can then be hollowed out and filled with other materials to make things such as tiny magnetic posts. But such well-ordered films tend to be too thin to mold useful magnetic posts. And in thicker films the cylinders point in random directions.

Russell's team set out to orient those cylinders in thick films. Starting with two-part polymers made from standard plastic precursors, polystyrene and polymethylmethacrylate (PMMA), they deposited a thick layer over a surface. The polymers self-segregated into randomly oriented PMMA cylinders in a polystyrene matrix. The researchers then simply exposed the film to an electric field oriented perpendicular to the surface. Because the lowest energy state for each cylinder is to follow the path of the lines of the external field, the cylinders stood at attention like an army of tiny soldiers.

To hollow out those cylinders, the researchers exposed their film to ultraviolet light, which breaks apart the PMMA and forges links between neighboring polystyrene molecules holding them in place. The result was a polystyrene matrix honeycombed with tiny tubes, which the team filled with cobalt. By changing the size of the copolymer components, “we can control the size of the cylinders [and nanowires] from 13 to 130 nanometers,” Russell says.

Although the researchers have yet to test the magnetic properties of individual cobalt posts, they say the array as a whole shows promising characteristics. One is unusually high coercivity, or magnetic resistance, a trait that suggests cobalt nanowire arrays may be better than standard magnetic media at holding their magnetization when subjected to heat or other random fluctuations.

Russell notes that the team is pushing ahead with other applications as well. Silicon-filled holes may be useful as tiny electron storage devices for other types of computer memory. Cylinders lined with particular compounds may form membranes that let certain chemicals pass through while blocking others. The researchers also are teaming up with other groups to use the holes to trap large membrane proteins, the better to decipher their three-dimensional structure. If even one of these uses pans out, plastic templates with nanosized pores could have a big future.

6. AGING RESEARCH

# Old Flies May Hold Secrets of Aging

1. Elizabeth Pennisi

Sensible people know better than to believe in pills that promise perpetual youth or weight loss without dieting. But then Stephen Helfand isn't known for always being sensible. Seventeen years ago, he left a lucrative career as a neurologist at a prominent Boston hospital to search for aging genes in fruit flies, a task many considered hopeless at the time. Yet, as he and his colleagues at the University of Connecticut Health Center in Farmington demonstrate on page 2137, there's lots to be learned by departing from the norm.

Helfand's team has discovered a gene that, when altered, can double the average life-span of fruit flies and may one day lead to that long-awaited miracle pill. “This [gene] provides optimism that it may, indeed, be possible to manipulate active life-span beyond the constraints that ordinarily apply in natural evolution,” says Seymour Benzer, the geneticist at the California Institute of Technology in Pasadena who in 1998 discovered a different fruit fly aging gene, dubbed methuselah (Science, 30 October 1998, p. 856).

Preliminary data suggest that the protein encoded by this gene, called Indy for “I'm not dead yet,” transports and recycles metabolic byproducts. Helfand thinks that defects in the gene, two copies of which exist in each fruit fly, can lead to production of a protein that renders metabolism less efficient; as a result the body functions as if the fruit fly were dieting, even though its eating habits are unchanged. As such, the discovery provides a “clear genetic link between metabolism and the rate of aging,” comments Tomas Prolla, a geneticist at the University of Wisconsin, Madison. The work may lead to a better understanding of how metabolism plays into aging. It may also illuminate why worms, fruit flies, and rodents, at least, live longer on a spartan diet.

Helfand came upon Indy by accident. In mutant strains of fruit flies produced for a different experiment, he and Connecticut's Blanka Rogina noticed that some lived longer than usual. Working with Connecticut's Robert Reenan, they first did some experiments to make sure that a mutation in Indy and not some other factor was indeed conferring longevity on these flies. It was. They determined that fruit flies carrying one good copy and one defective copy of Indy lived the longest, averaging 70 days versus the usual 37—“quite impressive extension of life-span,” notes Judith Campisi, a cell and molecular biologist at Lawrence Berkeley National Laboratory in California. With two altered copies of the gene, flies live only about 20% longer than the norm.

The Connecticut team tracked down other mutant strains with defects in the Indy gene, culling some from their own stocks and those of colleagues. They eventually came up with five different mutant versions of Indy, one copy of which always made the fruit fly live much longer. They chose the name based on a quip in the movie Monty Python and the Holy Grail.

In each mutant, Indy seems to extend life-span without exacting any other costs from the fruit fly. In tests, individuals belonging to the Indy strains flew just as well, ate just as much, and courted each other with as much vigor as did their shorter lived counterparts. They started reproducing at the same age, laid as many eggs, and even continued to reproduce long after other flies had stopped—suggesting they retained their youthful vitality.

Indy codes for a protein that resembles a sodium dicarboxylate cotransporter, a membrane protein found in many organisms, from bacteria to mammals, including humans. In mammals, dicarboxylate cotransporters show up in cells in the digestive tract, placenta, liver, kidney, and brain, where they transport metabolic intermediates across the cell membrane. In the fruit fly, the gene “is right at the places you'd like it to be” to serve a similar function, Helfand notes: It is active in fat bodies—which function as the liver in insects—the midgut, and in cells called oenocytes, which appear to store glycogen and be involved in metabolism. “Perhaps it's altering the nutrients, either their utilization or absorption, or making intermediate metabolism slightly less efficient,” he suggests. “Either way, it may be the genetic equivalent of caloric restriction.”

Given these data, the Indy mutants could prove a gold mine to aging researchers, says Prolla, as they may help clarify antiaging mechanisms. He'd like to see what happens to the life-span of flies that have both a mutation in Indy and one in genes that alter the fly's ability to deal with oxidative stress, another putative cause of aging. If those flies lived even longer, “it would suggest that there are in fact many routes for intervention in the aging process,” Prolla adds.

Despite their enthusiasm, aging researchers stress that much work remains. “The genetic evidence is strong, but there is some biochemistry that needs to be done to show that this really works the way they say it does,” says Campisi. Researchers have not studied cotransporters extensively in humans, except those in the kidney. So exactly how they work in, say, the gut of humans or flies remains unclear. Researchers must also find out how the protein works in different mutant strains.

But Helfand is hopeful. The Indy protein's location and nature suggest that eventually, “it may be possible to design a drug that can extend life,” he suggests. “The drug may very well work with weight control, too.” And that sounds like a chance even sensible people might be willing to take.

7. ASTROPHYSICS

# Young X-ray Satellite Rattles Old Ideas

1. Govert Schilling*
1. Govert Schilling is an astronomy writer in Utrecht, the Netherlands.

A year after its launch, Europe's XMM- Newton x-ray satellite is proving its mettle. Last week in Paris, scientists gave a sneak preview of findings that challenged cherished beliefs about what kind of material swarms around supermassive black holes in the middle of active galaxies, and how compact clusters of galaxies chill out. Those results and others, appearing next month as 56 papers in a special issue of the European journal Astronomy and Astrophysics, constitute what Roger Bonnet, director of science for the European Space Agency (ESA), describes as “monster astronomy.”

Since the mid-1980s, scientists have believed that x-rays they detected in active galactic nuclei were partly absorbed by clouds of warm gas far from the black hole. When XMM-Newton trained its spectrometers on two such nuclei, astronomers expected to see the hallmarks of such a gas shroud: x-ray spectra riddled with gaps where the atoms in the gas absorbed radiation passing through it. But XMM-Newton had other ideas.

“The spectra didn't show the spiky absorption signatures that we expected on the basis of the [warm-absorber] model,” says Masao Sako of Columbia University in New York City. Instead, as Sako was the first to realize, the spectra made much more sense if the spectral features originated from very hot gas orbiting extremely close to the black hole. What astronomers were seeing, Sako theorized, were not absorption lines but emission lines, distorted in ways that indicated that the gas must be moving close to the speed of light.

Material far from a black hole could not orbit so swiftly, says Jelle Kaastra of the Space Research Organization Netherlands (SRON) in Utrecht, the project scientist of the XMM-Newton spectrometers. That meant the gas had to be very close to the “edge” or event horizon of the black hole, just a few million kilometers from the center of the hole itself. The results also imply that the black hole is spinning, Kaastra says, because general relativity predicts that there are no stable orbits close to a stationary black hole. Physicists had long sought convincing evidence of such rotating black holescalled Kerr black holes, after the New Zealand physicist who predicted their properties mathematically. Now they have it.

“This is one of the outstanding results of XMM-Newton,” says Johan Bleeker of SRON. “For the first time, we're studying emission from the accretion disk itself, very close to the central, rotating black hole.” In principle, Bleeker says, observations like these should enable astronomers to derive the geometry, physical properties, dynamics, and chemical composition of the disk.

But Joachim Trümper of the Max Planck Institute for Extraterrestrial Physics in Garching, Germany, says discrepancies in the XMM-Newton data must be resolved before other physicists will accept them as conclusive. “This is certainly not the end for the warm-absorber model,” Trümper says.

XMM-Newton's second challenge to conventional wisdom targets compact clusters of galaxiesenormous congregations of hundreds or thousands of individual galaxies, generally dominated by a giant elliptical galaxy in the core. Older x-ray observations showed that the space between galaxies in clusters contains hot, x-ray emitting gas. Astronomers have long believed that the hot gas slowly flows into the cluster core, cooling as it goes, says Andrew Fabian of Cambridge University in the United Kingdom, who led one of the groups that proposed the model in 1977. Because the gas in the cluster's core is denser than gas farther out, Fabian explains, it cools more efficiently, radiating its energy away in x-rays. As a result, the pressure near the core drops. Hotter gas from the outskirts of the cluster then starts to flow inward, where it cools in turn.

If the cooling-flow model were correct, XMM-Newton should detect signs of cold gas in the inner parts of the clusters. In particular, astronomers say, its sensitive spectrometers should find the spectral signatures of moderately ionized iron atoms that have lost just a few of their electrons, an indicator of a low-energy environment. They don't. “The spectra provide us with a very significant [lower] limit on the temperature,” says Bleeker. “This puts the cooling-flow model in jeopardy.”

But Fabian thinks it's too early to write off the model. In a paper accepted for publication in the Monthly Notices of the Royal Astronomical Society, he describes five other possible explanations for XMM-Newton's results. Still, he admits the new observations have made things more complicated.

With so many intriguing clues emerging from just a few months' worth of observations, x-ray astronomers are confident that XMM-Newton will revolutionize the field. “This is not the end of the story,” Bonnet says. “I expect the observatory will continue to send back such interesting science results for the next 10 years.”

8. GENE THERAPY

# FDA Moves Against Penn Scientist

1. Gretchen Vogel

The U.S. Food and Drug Administration (FDA) has begun proceedings that could disqualify gene therapy researcher James Wilson of the University of Pennsylvania in Philadelphia from conducting any future clinical trials. Wilson, who is head of the university's Institute for Human Gene Therapy, oversaw the trial in which 18-year-old Jesse Gelsinger died after a genetically altered virus was injected into his liver (Science, 17 December 1999, p. 2244).

Disqualification is the harshest penalty the FDA can impose on an investigator. It bars a researcher from receiving drugs for use in clinical trials—in effect, preventing that investigator from administering experimental drugs to patients. In a 30 November letter to Wilson, the FDA stated that Wilson had “repeatedly or deliberately violated regulations governing the proper conduct of clinical studies.” The agency wrote that Wilson and his colleagues enrolled patients who were ineligible for the trial, did not monitor patients properly, did not halt the trial when patients experienced serious side effects, and failed to inform patients that a trial of a similar drug had severely sickened monkeys. The FDA has also issued warning letters to two of Wilson's collaborators in the study—Steven Raper of the Institute for Human Gene Therapy and Mark Batshaw of Children's National Medical Center in Washington, D.C.

“This is obviously a very serious matter,” the university said in a brief statement. “We know that Dr. Wilson understands its importance, is reviewing the letter carefully and intends to respond in a timely way.” Wilson has 30 days to reply to the FDA's letter. After reviewing Wilson's response, FDA administrators will make a final decision.

This is a “drastic” step, says Inder Verma of the Salk Institute for Biological Studies in La Jolla, California, who headed a special working group at the National Institutes of Health that investigated the Gelsinger trial. But Savio Woo, a gene therapy researcher at Mount Sinai School of Medicine in New York City and past president of the American Society of Gene Therapy, says that vigorous FDA oversight will strengthen gene therapy research.

9. NEUROSCIENCE

# Immune Molecules Prune Neural Links

1. Laura Helmuth

The developing brain starts off as a tangle of neuronal connections, then activity reinforces some of these connections and causes others to atrophy. Over the past few years, neurobiologists have been on the prowl for molecules that help the nervous system make and break these connections, and they've come up with a few contenders. But now Carla Shatz's team at Harvard Medical School in Boston is proposing an unlikely new candidate: a type of protein previously known for its role in helping the immune system fend off viruses and other foreign invaders.

In work described on page 2155, Shatz and her colleagues suggest that the class I major histocompatibility complex (MHC) proteins are necessary for the formation of normal neuronal connections in a visual area of the brain during development. Later in life, they're called into play in the hippocampus, a brain area involved in memory and learning. The work shows “a completely unexpected function for the molecules,” says neurobiologist Marc Tessier-Lavigne of the University of California, San Francisco.

The current results are an outgrowth of observations that Shatz's team, then at the University of California, Berkeley, reported 2 years ago. While examining how gene expression patterns in the brain react to changes in retinal activity, the researchers found, to their surprise, that the genes encoding class I MHC proteins are active in the developing brain. To explore what these seemingly alien molecules were doing there, Gene Huh, a postdoc in the Shatz lab, turned to three strains of mice that had been genetically altered. Two of these strains lacked the ability to display class I MHC proteins in their normal location on the cell surface; the third lacked part of a receptor that T cells use to respond to class I MHC proteins. In the first phase of their work, Huh and his colleagues tested how these gene knockouts affected the development of the animals' visual systems.

Like many animals, mice aren't born with the ability to see. The visual system matures only when neural signals, originating either from spontaneous neural firing in the retina before the eyes open or from looking around afterward, help organize the parts of the brain that receive and process visual information. Huh found that this process, which involves both strengthening frequently used connections and pruning useless ones, is disrupted in all three knockout mice.

The visual signal's first stop after the eye is the lateral geniculate nucleus (LGN). There, neuronal projections from the retina normally form what Huh describes as a “big misshapen doughnut.” The doughnut itself, occupying most of the LGN, receives input from the eye on the opposite side of the body, while a small “doughnut hole” in the middle of the LGN gets projections from the same-side eye. But in the knockout mice the doughnut hole is much larger, implying that the inputs from the two eyes overlap. The finding suggests that in the absence of functioning class I MHC proteins, the normal pruning of connections that should have occurred in the LGN is defective.

Similar strengthening and weakening of neuronal connections is thought to occur during memory formation in the adult hippocampus. So Shatz and her colleagues next looked at synapses in that brain region, where their earlier work had shown that the class I MHC genes are also active.

When postdoc Lisa Boulanger stimulated the hippocampal neurons, she found that neurons in the knockout mice reacted strangely. Long-term potentiation (LTP), the strengthening of signals with stimulation, was enhanced: Hippocampal neurons in the knockout mice responded more dramatically than did those in normal mice to a high-frequency stimulus that can evoke LTP. And when she applied low-frequency stimulation, which causes the synapse weakening known as long-term depression, neurons in the knockout mice failed to rein in their signals as they should have. “To us,” says Shatz, “the results imply that there may be a commonality of cellular and molecular mechanisms” in how neurons in the hippocampus and the developing nervous system respond to activity.

Huh and Shatz suspect that the class I MHC proteins also help neurons tune their connections in other areas of the brain. The mouse carries about 30 different varieties of the protein. When Huh tested where four of the genes are expressed, he found that different ones are active in different places in the brain (see figure), possibly tailoring nearby neurons to fit into the correct neural pathways.

Despite the evidence indicating that the class I MHC proteins are somehow involved in refining neuronal connections in the brain, researchers don't yet know how the molecules are acting. Immunologist Hidde Ploegh of Harvard says he's “intrigued by the possible role of [these molecules] in something that appears to have no immunological correlate.” But he awaits the details of how they might be helping the nervous system figure out which connections should atrophy.

10. NEUROPSYCHOLOGY

# Language Affects Sound Perception

1. Charles Seife

NEWPORT BEACH, CALIFORNIA— Neuropsychologists may owe a debt to the devil. At the 140th meeting of the Acoustical Society of America here last week, University of California, San Diego, psychologist Diana Deutsch demonstrated that an auditory illusion based on the tritone—also known as the “devil in music”—is perceived differently by listeners with different linguistic histories. And those perceptions might help psychologists understand how the brain rewires itself during childhood.

Played together, two notes a half-octave apart (such as C and F sharp) sound jarring; medieval musicians considered this combination, a tritone, so discordant that they dubbed it the “diabolus in musica.” But a tritone is music to Deutsch's ears. For more than a decade, she has been studying an auditory illusion—the acoustical equivalent of an optical illusion—based upon the tritone.

With a computer, Deutsch created ambiguous notes by superimposing tones from many octaves and carefully shaping the relative loudness of the higher and lower frequency components, masking how high or low the note is. Although listeners can perceive one of these notes as, say, a C, they can't tell its octave, whether high C, middle C, or low C. Indeed, the tone doesn't really belong to any octave at all.

Things get interesting when people compare tritone pairs of ambiguous notes, such as an ambiguous C with an ambiguous F sharp. Even though neither note is higher or lower than the other—because higher and lower don't have any meaning with ambiguous notes—people consistently perceive one tone as high and the other as low. But strangely, they don't agree which is which. “The musical illusion is perceived very differently by different people,” says Deutsch. This is the tritone paradox.

Things got even weirder when Deutsch played ambiguous tritones to different groups of people. In 1992, she noticed that people from California and people from Southern England hear tritones in the opposite way; if a Californian thinks that a C is above an F sharp, the Britisher will swear that the F sharp is higher than the C. This led some psychologists to believe that a person's perception of ambiguous tritones depends strongly upon the intonations of the language he learns as a child. Since then, psychologists have been trying to prove it.

At the Acoustical Society of America's meeting last week, Deutsch described her latest experiment, in which she played ambiguous tritones to two groups of subjects who had emigrated to California from Vietnam. The first group came to the United States as children, and although Vietnamese was their first language, most no longer spoke it fluently. The second group, on the other hand, arrived in the United States as adults and spoke little English. The two groups perceived the tritones in the same way—but differently from their California neighbors. “This study presents strong evidence, we believe, that individual differences [in perceiving the tritones] are caused by individual speech patterns to which [the subjects] are exposed early in infancy,” says Deutsch.

“It reinforces the idea that early linguistic background affects perception,” says Magdalene Chalikia, a psychologist at Minnesota State University in Moorhead, who has shown that Greek speakers and English speakers perceive tritones differently. The tritone paradox gives neuropsychologists intriguing hints about the effects of training on the brain. The results suggest that as an infant learns its first language, the brain may adjust its neural connections in a way that affects the perception of sounds. But for the moment, scientists have little idea which languages cause which interpretation of the tritone paradox, much less how each language rewires the brain differently. “You can't make predictions,” sighs Chalikia. “It's frustrating.” The devil, it turns out, is in the details.

# California Sets Up Three New Institutes

1. Evelyn Strauss

Three University of California (UC) campuses were chosen last week as sites for a new $900 million program designed to keep the state a world leader in research and to bolster its economy. Each of the three schools will receive$25 million a year for 4 years from the state, with companies and other sources putting up at least twice that amount.

The money will create California Institutes for Science and Innovation at the Los Angeles (UCLA), San Francisco (UCSF), and San Diego (UCSD) campuses. UCLA will team up with UC Santa Barbara on a nanosystems institute that will be led by Martha Krebs, former director of the U.S. Department of Energy's Office of Science. UC Berkeley and UC Santa Cruz will join with UCSF in an institute on bioengineering, biotechnology, and quantitative biomedical research headed by David Agard, a UCSF professor of biochemistry and biophysics. The third institute, on telecommunications and information technology, will be a collaboration between UCSD and UC Irvine led by UCSD computer science and engineering professor Larry Smarr.

View this table:

“We'd like this to be a magnet for the best and brightest of the scientific community,” says California Governor Gray Davis, who pushed the idea through the state legislature (Science, 26 May, p. 1311). “We can't make them come, but we'd like them to know they're welcome.” Davis also promised to lobby next year for a fourth center, based at Berkeley, that would apply information technology to critical societal problems such as transportation, education, emergency preparedness, and health care. The Berkeley proposal fell just short in a competition among six finalists.

The contestants were encouraged to dream up novel collaborations and projects, and the winners were eager to describe how their research plans will push the boundaries of their field. “The growth of the wireless Internet will lead to radical change,” says Smarr, describing sensors embedded in bridges, cars, and even people that may someday transmit information to a computer miles away that can assess problems such as stresses during an earthquake or wear-and-tear on a vehicle's brakes. “Wouldn't it be nice if you got a call on your cell phone that said, ‘Hello, we thought you'd like to know that your right front brake will fail in about 100 miles.’”

Officials also emphasized that the institutes should tackle topics not historically addressed on their own campus. Developing innovations in engineering and technology, says Agard of UCSF, a campus devoted to the health sciences, is “a new game in biology and [requires] resources that go beyond what normal medical schools can come up with.”

Although winning entrants were restricted to the 10 UC campuses, the new institutes are also hoping to work with some of the state's most prestigious private schools, including Stanford and the California Institute of Technology. “I envisioned a research process open to the best minds, wherever they were,” says Davis. The same goes for collaboration with industry, which is expected to contribute heavily to the new institutes. “Places like Hewlett-Packard and Sun want our students,” says Krebs, “and they also want access to the results of our research.”

12. ENVIRONMENTAL TOXICOLOGY

# Treaty Takes a POP at the Dirty Dozen

1. Jocelyn Kaiser,
2. Martin Enserink

Last month's talks to mitigate global warming may have flopped, but this week brought some consolation to those concerned about the planet's environmental health: the first-ever global agreement to abolish a class of dangerous industrial chemicals. The treaty, finalized by representatives of 122 countries meeting in Johannesburg, South Africa, also spells out a process to determine the next chemicals to be proscribed.

The treaty on persistent organic pollutants, or POPs, as they're known, will ban or phase out 12 long-lived pesticides and other toxic chemicals still used in many developing countries. These chemicals are slow to break down, and they accumulate in body fat. Concern about the toxicity of POPs “goes all the way back to Rachel Carson's” warning about DDT and eggshell thinning in birds, notes reproductive biologist Louis Guillette of the University of Florida, Gainesville. To the relief of some public health experts, the treaty permits one exemption: the limited use of DDT to control malaria, a practice still common in Africa, Latin America, and Asia. Government officials, environmental groups, and malaria researchers who had wrangled over the treaty for 2 years pronounced themselves satisfied with the results.

The POPs treaty was organized by the United Nations Environment Program to address the “dirty dozen”. These substances get carried by global weather patterns to the polar regions, where they've been blamed for a variety of problems in wildlife and people. High levels of polychlorinated biphenyls (PCBs) in the breast milk of Inuit women, for example, have raised concerns about possible immunological and intellectual deficits in children. And according to one controversial theory, trace levels of POPs acting as “endocrine disrupters” may contribute to problems such as lower sperm counts and cancer in the general population.

The United States and most other developed countries banned PCBs and most POP pesticides years ago. However, they're still widely used in places such as India and Latin America. The treaty finalized last week would ban eight of these pesticides immediately. Two industrial byproducts on the list, dioxins and furans, will be reduced right away and eventually eliminated “where feasible,” for example, by clamping down on open trash burning. PCBs, used in electrical transformers, will be allowed until 2025 as long as equipment is maintained to prevent leaks. To help developing countries destroy stockpiles and develop alternatives, delegates also agreed to an estimated $150 million annual fund run by the U.N.'s Global Environment Facility. The treaty will be signed in May in Stockholm and goes into effect once 50 countries have ratified it. Ironically, the new treaty makes an exception for the most well-known and infamous of the “dirty dozen”: DDT. After intense pressure from a group of malaria experts, who argued that there was no effective substitute, the convention decided to allow the limited use of DDT for mosquito control. Roger Bates of Africa Fighting Malaria, a loose coalition of DDT supporters in South Africa, says that banning DDT now would be like “crossing a street with heavy traffic to avoid a crack in the pavement.” His group gained support earlier this year from the World Health Organization. Despite that exemption, Bates says the treaty “is not 100% wonderful.” Its tone may embolden donor agencies to pressure developing countries to abandon DDT, he says, adding that it is already more difficult to come by. Under the treaty, countries must also report their DDT use in a special register, raising its cost and revealing its users. “All the green groups in the world are going to know who is using DDT and where,” he says. Another major sticking point were the rules for adding chemicals to the list. European countries and environmentalists argued for incorporation of the “precautionary approach,” which says that it may be necessary to take action against an environmental threat even when the scientific evidence is incomplete. But U.S. officials and the chemical industry worried that such an approach would ignore risk analysis, which bans chemicals only if enough data show that they're dangerous. In the end, delegates compromised by explaining that precaution would be “an integral part of—and not separate from—the overall scientific process,” according to a statement from U.S. State Department negotiator Brooks Yeager. According to Kip Howlett of the American Chemistry Council, which represents chemical manufacturers, a key step was to reference a definition agreed on at the 1992 Rio Earth Summit, which requires cost-benefit analysis. The rules preclude a ban based on persistence and bioaccumulation alone, Howlett says, an approach that is gaining ground in Europe (Science, 1 December, p. 1663). That interpretation will be put to the test in analyzing the next round of chemicals. Scientists say that there are a number of obvious candidates, such as the pesticide dicophol and perfluorooctane sulfonate. The latter was used in Scotchgard fabric protector products until 3M pulled them off the market in May even without evidence of harmful effects in people. Another category is fire retardants containing brominated compounds that are similar to PCBs. Some animal studies have shown that they affect the thyroid hormone, but “data are really limited,” says Linda Birnbaum, a dioxin toxicologist at the U.S. Environmental Protection Agency. However, concentrations have been rising for two decades in women's breast milk in Sweden, and more recent data from a New York State study not yet released will show “levels that are really high,” Birnbaum says. If the POPs treaty can finger such poisons and lower their concentrations, it may rank with efforts to curb ozone-destroying fluorocarbons as a major environmental health triumph. Like those pollutants, POPs “have consequences in parts of the globe far from where they're released,” Guillette says. “The treaty is arguing the same kind of thing: Somebody has to take responsibility.” 13. SEQUENCE # Plants Join the Genome Sequencing Bandwagon 1. Elizabeth Pennisi The complete genome sequence of the first higher plant is revealing unexpected similarities to—and differences from—the genomes of other organisms, including humans COLD SPRING HARBOR, NEW YORK— In simultaneous press conferences across the globe, an international consortium announced this week that it has finished the first genome sequence of a higher plant. For plant biologists, the eagerly awaited genome of this small weed, Arabidopsis thaliana, offers a window into the genetic makeup of all plants, including key crops. And it's a clear window indeed, as the six international sequencing teams on three continents have pulled off a coup: They produced a genome sequence that is more accurate than that of any multicellular organism which has been published to date. Through this window, they are seeing for the first time that plants may be much more complex than many biologists have imagined. And that realization turns out to be a bit humbling. With its estimated 125-million-base sequence, Arabidopsis has a surprisingly high number of predicted genes—some 25,500, compared with 13,600 in the fruit fly. That total—which exceeds by several thousand that of any other genome yet analyzed—comes impressively close to low-end estimates of the number of human genes (Science, 19 May, p. 1146). “It says that you have to have more respect for plants,” concedes Nobel laureate James Watson, co-discoverer of the double helical structure of DNA. “Plants may be sessile, but they are not stupid,” adds geneticist Michael Sussman of the University of Wisconsin, Madison. Both were on hand this week at Cold Spring Harbor Laboratory (CSHL) along with more than 150 Arabidopsis researchers for a gathering that was part science and part celebration. The assembled researchers were giddy with pride and anticipation. For them, the sequence, which is described in the 14 December issue of Nature, signals the dawn of a new era. “This is a very special time in the whole history of plants,” says Daniel Cosgrove, president of the American Society of Plant Physiologists. He expects that the sequence data will speed the identification of genes to improve agriculture as well as expand knowledge of basic plant biology. Not bad for a plant that 20 years ago was considered decidedly offbeat (Science, 6 October, p. 32). Even a decade ago, when Arabidopsis enthusiasts in the United States and Europe proposed understanding the genome of this mustardlike weed, many of their colleagues had to be convinced. Little happened until 1994, when, with encouragement from the European Commission, several European labs joined forces and began a pilot sequencing project. Coordinated by Michael Bevan, a geneticist at the John Innes Centre in Norwich, United Kingdom, the project quickly swelled to 17 European groups. It went transatlantic in 1996, when The Institute for Genomic Research (TIGR) in Rockville, Maryland, signed on; CSHL teamed up with Washington University; and Stanford, the University of Pennsylvania, and the U.S. Department of Agriculture lab at the University of California, Berkeley, formed a third group. Satoshi Tabata and his team at the Kazusa DNA Research Institute in Kisarazu, Japan, also joined the consortium. Their efforts have yielded 115 million bases of Arabidopsis DNA that cover the five chromosomes. These sequencers also ventured into the sequence and structure of the centromere, an important but rather inscrutable region of the chromosome that assists in chromosome pairing (see p. 2057). To date, chromosomes 1 and 5 have about three gaps each left to fill, far fewer, for instance, than the 1200 gaps in the Drosophila genome sequence. “It's by far the most high-fidelity multicellular eukaryotic genome to date,” says Jeffrey Dangl, a molecular geneticist at the University of North Carolina, Chapel Hill. Although each sequencing group analyzed the sequence as they produced it, two groups in particular scrutinized the data intensely, using computers and the human eye to pick out all the genes and their coding regions. TIGR and the Munich Information Center for Protein Sequences in Germany found that it has lots of duplicated genes. They estimate that, overall, duplicated regions make up 58% of the genome, likely because of whole-genome duplications that occurred 100 million years ago. They also assigned tentative functions to roughly 70% of the genes, in part by comparing the genes and their proteins to those of other organisms already in the genome databases. From these comparisons other surprises have emerged. For one, some 100 Arabidopsis genes have counterparts in humans that cause disease, including the genes involved in cystic fibrosis and breast cancer. This plant also has a relatively large number of genes for water-channel proteins, as well as many more protein kinase genes than expected. Because protein kinases are involved in cell signaling pathways, “there are an awful lot of ways Arabidopsis cells are talking to each other and getting information from outside the cells,” says Elliot Meyerowitz, a geneticist at the California Institute of Technology in Pasadena. Another intriguing find is that the set of genes for cell-to-cell communication varies dramatically between Arabidopsis and humans. By contrast, many genes involved in basic cell functions, such as cell division, tend to be conserved among species—in other words, a yeast gene might work just as well in Arabidopsis—suggesting that the genes existed in a common ancestor of all organisms. But not so the cell communication genes. “[That] fits with the idea that multicellularity evolved separately [in plants and animals],” says CSHL's Robert Martienssen, a plant geneticist. These are the first of many insights expected to pour out of the genome, not just for learning about Arabidopsis but also about other plants. With this genome in hand, plant biologists should be able to find key genes in other plants far more easily. As Bevan points out, such insights may lead to the development of crops better suited to developing countries, plants designed to soak up more carbon dioxide, or other applications that cannot yet be imagined. For that reason, he predicts, the genome sequence from this tiny plant “will have as much impact as the human genome.” 14. ECOLOGY # Stalking the Wild Mustard 1. Elizabeth Pennisi Arabidopsis thaliana is not just for molecular biologists anymore. A few intrepid ecologists and evolutionary biologists are now touting its merits Several years into his Ph.D. studies, Massimo Pigliucci realized he had a problem. He wanted to study how the environment interacts with a species' genetic makeup over time, and he had been focusing on Lobelia, a tall, native perennial much studied by ecologists. But Lobelia “was hard to use,” he recalls, because little was known about its genetics. So Pigliucci shifted to a small weed typically studied by molecular and developmental biologists rather than ecologists. His committee at the University of Connecticut was appalled. The weed, Arabidopsis thaliana, “wasn't a real plant” as far as his advisers were concerned. And, they asserted, it had no intriguing features worthy of ecological or evolutionary investigation. A decade later, Pigliucci, now at the University of Tennessee, Knoxville, and a small but growing number of other plant ecologists and evolutionary biologists are proving those advisers wrong. They have searched the globe for natural variants of this weed and of its relatives, which will enable them to probe the evolutionary history of this plant (see sidebar). Bucking the trend among field biologists to focus on wild flora in pristine places, these researchers are capitalizing on decades of intensive research on Arabidopsis. Thousands of research papers have already examined the physiology, development, biochemistry, and genetics of this plant. Indeed, this week the full genome sequence of Arabidopsis is being published—the first complete sequence of any plant (see Nature, 14 December, p. 796, and this issue of Science, p. 2105). Now it's time for more ecologists and evolutionary biologists to make use of these data, says Thomas Mitchell-Olds, an ecologist-turned-quantitative geneticist at the Max Planck Institute of Chemical Ecology in Jena, Germany. For instance, he and others want to tease out how genetic variation helps Arabidopsis thrive under different conditions. Even mainstream ecologists are beginning to see the potential of this lowly weed. Stanford plant ecologist Harold Mooney, for one, is convinced that Arabidopsis will help evolutionary biologists and ecologists answer numerous other questions, and he is encouraging his colleagues to take a chance on this unlikely plant. Specifically, says Mooney, Arabidopsis offers “an incredible opportunity” for them to understand the genetic basis of the traits they study. Until now, researchers have had trouble connecting the genotype—an organism's particular mix of genes—with its phenotype—how that organism develops and acts. But with Arabidopsis, “you can precisely answer this question, because it's got great genetics,” says Michael Purugganan, an evolutionary geneticist at North Carolina State University in Raleigh. Those great genetics include not only the sequence but also mutants that Arabidopsis biologists have developed for almost every gene. By growing these mutants—or natural variants—under different conditions, Pigliucci and other plant ecologists are homing in on the genes responsible for specific traits, such as the timing of flowering or the production of protective compounds. The sequence information “allows us to plan our experiments more carefully and analyze our data more thoroughly,” says Naohiko Miyashita, a population geneticist at Kyoto University, who switched from Drosophila to Arabidopsis 6 years ago. Rodney Mauricio, an ecological geneticist at the University of Georgia, Athens, is using Arabidopsis to look at how insect pests affect the evolution of plants. Although he complains about how delicate this weed is—for his next experiment he recently planted 4800 seedlings, each no bigger than a thumbtack—he's one of the pioneers in using Arabidopsis in the field. Over the past decade he and his colleagues have established that insect pests prompt plants to develop tiny hairs called trichomes and that these have a “cost”: The hairier plants produce fewer seeds. “It's a fairly important adaptive story,” Mauricio notes. Now he's tracking down the genes responsible for trichome growth and density. Says Mauricio, “The idea is to move toward a complete understanding of evolution from the ecological side to the molecular side.” Another convert is Johanna Schmitt of Brown University in Providence, Rhode Island. She got hooked on Arabidopsis when Pigliucci joined her lab as a postdoc 6 years ago. Together, they began to study how the weed reacts to crowding by other plants. When the plant's light-sensing pigments, called phytochromes, are shaded by close neighbors, they somehow trigger rapid growth by the main stem, causing the plant to become tall and gangly and prompting early flowering. Molecular biologists had already tracked down the pigments and their genes; Schmitt, Pigliucci, and colleagues used the Arabidopsis strains developed for those molecular studies in their work. By planting the various strains, each of which lacks a gene for a particular pigment, under different crowding conditions, “we were able to measure natural selection on shade avoidance,” Pigliucci says. Purugganan has teamed up with Schmitt's group to track down the genes that regulate when Arabidopsis flowers, taking advantage of natural variation in North American populations. In North Carolina, Arabidopsis seeds germinate in the fall, forming a small rosette of leaves that hugs the soil and waits out the winter before sending up a stalk and, eventually, flowering. But in harsher climates, nothing happens until spring, when the seeds sprout immediately and flower a few weeks later. In other places, such as Rhode Island (in southeastern New England), Schmitt and her postdocs Lisa Dorn and Cynthia Weinig find stands of Arabidopsis where both growth patterns exist. To understand the variation in this trait, Weinig and her colleagues from Brown and North Carolina last year planted inbred lines of Arabidopsis that differed in particular genes, such as those coding for proteins that sense day length, both in Rhode Island and in North Carolina. They also planted Arabidopsis from Rhode Island in North Carolina and vice versa. Over the ensuing months the researchers monitored when these plants sprouted and when they bloomed. They also noted traits such as the number of branches to see how these plants responded to the conditions in each place. Already, intriguing hints are emerging. The preliminary data suggest that, depending on the season, different genes are affected. The team's goal, Schmitt says, is to understand “what are the causes of selection on particular traits in [the] wild, what are the genes underlying individual traits, and how is selection acting on these genes.” Still others are studying how Arabidopsis and its bacterial or fungal pathogens coevolve. Ecological geneticist Joy Bergelson of the University of Chicago, for instance, used sequence data to study the R genes that help the plant recognize bacterial pathogens. Next, her group obtained Arabidopsis from various sites around the world and assessed their R-gene makeup. Contrary to what they expected, Bergelson and her colleagues have found that different versions, or alleles, of R genes have persisted in these populations for millions of years. That finding, along with other data from Mitchell-Olds's group, indicates that defense-related genes are not in an evolutionary arms race with pathogens, as current dogma predicts. Instead, suggests Mitchell-Olds, the frequency of the different alleles varies from year to year and population to population, enabling the plant to maintain its defenses against numerous threats. These insights represent just the beginning of what Mitchell-Olds and others expect to gain by moving Arabidopsis studies into the field. “Ecologists may like to work in wonderful pristine places,” Mitchell-Olds says, “but this is a system where we can get answers.” 15. # Arabidopsis Kin Help Keep Genetics Studies All in the Family 1. Elizabeth Pennisi An unobtrusive weed often found at roadsides and other disturbed sites has risen from its humble roots to become the model organism, or “lab rat,” for studying plant molecular biology. Now, the far-flung relatives of that weed, Arabidopsis thaliana, are slowly becoming celebrities as well. By comparing the genetic makeup and characteristics of A. thaliana to its cousins, evolutionary biologists can discern how the plant has changed through time. Before doing so, researchers first needed to clarify Arabidopsis's family ties. Over the past few years, teams led by Marcus Koch of the University of Agricultural Science in Vienna and Thomas Mitchell-Olds, a quantitative geneticist at the Max Planck Institute of Chemical Ecology in Jena, Germany, have done so. Working with evolutionary biologist John Bishop, now at Washington State University in Vancouver, they analyzed ribosomal DNA sequences and various gene sequences from well-studied Arabidopsis species and from species in closely related genera, such as Arabis, Cardamine, and Capsella. This work, reported in several journals in 1999 and 2000, has led to a new family tree: Arabidopsis now has 10 species, with A. lyrata and A. halleri being the most closely related to the “lab rat.” But there's a problem in studying these relatives: The molecular tools developed for A. thaliana don't necessarily work on them. Mitchell-Olds and researchers at six other European labs have spent the past 2 years adapting these tools for some of the kin. The researchers have cut the related genomes into malleable chunks and cloned them. They are also developing genetic maps, essentially a series of DNA landmarks along the chromosomes, of several relatives to enable investigators to track down genes more easily. Others are making sure that techniques for introducing or altering genes in A. thaliana work in these cousins as well. The ultimate goal, says Mitchell-Olds, is to provide plant biologists, especially evolutionary biologists, with the molecular and genetic equipment necessary for analyzing species that split off from the A. thaliana line 5 million, 10 million, and 30 million years ago. “We're hoping that researchers who want to make a close or distant comparison can get the stuff they need from a stock center,” several of which already exist, says Mitchell-Olds. Already, these relatives are proving their worth. One drawback to A. thaliana—at least in some eyes—is that it is a selfing plant that fertilizes its flowers with its own pollen, thereby creating stands of identical plants. A. thaliana's closest cousin, A. lyrata, provides a great alternative. Although it looks much like A. thaliana, it is not selfing; instead it has large flowers that attract insects carrying pollen from other plants. Another advantage is that, unlike A. thaliana, A. lyrata is not an invasive weed—in other words, it stays put. That means A. lyrata has had millions of years to coevolve with the local pests, whereas A. thaliana largely hasn't. Mitchell-Olds and Bishop have used A. lyrata and other kin to examine the evolutionary history of a family of antifungal proteins called chitinases. Meanwhile, at the University of Oulu in Finland, Outi Savolainen, who coordinates the European consortium on Arabidopsis evolutionary genetics, has been looking at various populations of A. lyrata to probe the genetics of flowering time in this species. This work should complement research in the United States on flowering times in A. thaliana (see main text). And these are just the beginning. Arabidopsis's sister genus Arabis, for example, contains a species that fungal pathogens infect. After infection, the pathogens force the plant to produce an orange pseudoflower—orange because it's covered with spores that are then carried away by pollinators deceived by the fake blossom. “We may be able to apply Arabidopsis information to eventually understand how that pathogen takes over that plant,” Bishop suggests. In short, notes Charles Langley, a population geneticist at the University of California, Davis, “there's a lot of cool stuff that can be done.” 16. CENTROMERES # A Journey to the Center of the Chromosome 1. Christine Mlot* 1. Christine Mlot is a science writer in Madison, Wisconsin. The Arabidopsis genome project is the first to give a detailed picture of the centromeres in a higher eukaryote As a postdoc at Stanford University in 1994, Daphne Preuss was examining mutagenized pollen grains under a microscope when she saw it: Amid all the dots of lone pollen, four grains were stuck together, tracing the shape of a tetrahedron. Having written her Ph.D. thesis on yeast, where such tetrads are standard and have been the foundation for its genetic analysis, Preuss knew she was looking at something powerful. “I immediately knew this [mutant] was the key to doing all kinds of genetic analysis” in plants, she recalls. “Life would be different.” That chance finding launched her career as a plant biologist. Some 6 months after she found it, an electron micrograph of the mutant pollen was on the cover of Science (3 June 1994), and Preuss was soon on her way to the University of Chicago, where she directs a lab that runs in large part on the power of her mutant find, dubbed quartet. The lemon-yellow pollen grains in which she spotted quartet were from the mustard plant Arabidopsis thaliana. What was unusual was that the four gametes were joined. Typically during meiosis in a plant or animal, the two chromosomes within a cell join; recombine, or exchange genetic material; then divide and separate twice into four haploid cells—the gametes. Each gamete, whether pollen or sperm, contains half the genetic complement. But in this newfound Arabidopsismutant, the standard diploid cell produces four adjoined haploid cells—a tetrad, as in yeast. By analyzing these four cells instead of random gametes, geneticists can chart recombination events with unprecedented precision. Preuss realized that this four-in-one mutant could reveal what happens during meiosis in plants as it had in yeast. It would also enable her to define the centromeres, which have been defined in yeast but which remain a black box in plants and animals. The centromere is a crucial stretch of DNA buried in the knotty terrain at the center of the chromosome. It plays a key role in meiosis, pairing up parental chromosomes and hitching them to protein motors that pull the chromosomes apart before cells divide. The dense, central region of the chromosome containing the centromere is readily visible under a microscope. Yet only in yeast have researchers been able to identify the exact DNA sequence of the centromere. Using the Arabidopsis tetrad mutants, Preuss has established where the centromeric region starts and stops on each of the five chromosomes, a first for a complex eukaryote. Now, by building “minichromosomes,” she and her colleagues are on their way to pinpointing where and how in that region the proteins attach in meiosis. The research is “blazing trails,” says Kelly Dawe of the University of Georgia, Athens, who is developing such minichromosomes in maize. Preuss's lab and quartet have also been indispensable to the Arabidopsis genome sequencing project, started in 1996 with the goal of deciphering the plant's 120-million-base-pair sequence. The fine-scale genetic map her group developed by using pollen tetrads boosted the unprecedented accuracy and completeness of the sequence of this model organism. Not only did the map enable Preuss to define the centromeric region, but it also enabled the six sequencing groups to assign unknown fragments of DNA, especially from the centromeric region, to their rightful places on the chromosomes. “We wouldn't have been able to have done it without her and [postdoc] Greg Copenhaver,” says W. Richard McCombie of the sequencing group at Cold Spring Harbor Laboratory in New York. ## Into the chromosome centers Under a microscope, the cinched region of the centromere is easily one of the most distinguishing features of the threadlike chromosomes. Indeed, cytologists captured the first images of the centromere in the late 1800s. But for all their centrality—literally and figuratively—in dividing cells, centromeres have escaped much dissection. “The centromere has been remarkably elusive to pin down,” says Brian Charlesworth, an evolutionary biologist at the University of Edinburgh in the United Kingdom. For decades both cell biologists and geneticists have attacked the problem of how chromosomes segregate and the role the centromere plays in this process. By the 1940s geneticists were mapping the locations of the dense centromeric regions in yeast and other fungi. This was possible because these organisms packaged their gametes in tetrads, which enabled geneticists to track recombination among the four cells. The centromeres always stayed put on the chromosomes they started out with, in contrast to other recognizable markers. Thus, the centromeres became the baseline for measuring distance to markers on the chromosome arms. From that point on, tetrad analysis became a stock in trade for geneticists, especially as yeast became the favored model system. In 1980, using tetrad analysis, researchers narrowed the location of the centromere to about a 4000-base stretch of DNA in budding yeast, the first organism to achieve that landmark at that level of resolution. Eventually, as molecular technologies improved, yeast's functional centromere—the precise bases involved in hitching chromosomes to the proteins—was whittled down to a 125-base stretch. Meanwhile, during the 1960s and '70s, cell biologists had developed staining and other techniques that visually highlighted the centromeric region on the larger chromosomes of plant and animal cells. These techniques also highlighted the corresponding protein structure the centromere meshes with. By that time, researchers were beginning to realize that eukaryotes had devised many ways to segregate their chromosomes—in other words, one centromere did not fit all. But except in yeast, the exact DNA sequence of the centromeres remained elusive, escaping detection even in the massive projects to sequence the genomes of fruit flies and humans. The problem is that the chromosome centers are jungles of difficult loops and repetitions, or heterochromatin, in contrast to the smooth runs of readable DNA, or euchromatin, on the chromosome arms. As sequencers—who decipher the genetic code of short fragments and then reassemble them in correct order—approach the central regions of the chromosomes, they run into long stretches of nothing but repeated bases, such as ATAT … AT. These stretches of repetitive DNA are next to impossible to piece together, as they contain few landmarks to orient them. The centromeric region is “the part of the genome that's ignored when it comes to estimating time to ‘complete’ genome projects,” says Gary Karpen, who studies the centromere in Drosophila at the Salk Institute for Biological Studies in La Jolla, California. In Arabidopsis, Preuss and her co-workers were able to find the right place for these “orphaned” stretches by analzying the results of 1000 crosses using the mutant pollen, each cross producing four plants, one from each cell of the tetrad. With tetrads, “we can literally redraw what happened in meiosis,” she says. They were able to find recognizable markers linked to these orphan stretches and then examine how often the markers stayed put with the centromeres in the tetrads generated in meiosis. The more often a marker separated from its centromere, the farther down the chromosome arm it was. The more often it stayed with its centromere, the closer in it was. The frequency of these separations, or recombinations, within a foursome was used to calculate distance between markers and centromeres. In this fashion the researchers generated successively finer scale maps of the genome, starting out with roughly one marker for every three megabases and finishing in the region of the centromere with one marker for every 10 kilobases. The frequency of recombination between a marker and its centromere also enabled the Preuss group to delineate where on the map the centromeric region starts and stops on all five chromosomes (Science, 24 December 1999, p. 2468). When they found no pattern of crossing over in the tetrads, the researchers knew that they had found markers in the genomic terra incognita of the centromeric region itself. Among the five chromosomes, these genetically defined centromeric regions vary in length from 1.4 megabases to 1.9 megabases, says Preuss. That's about 7% of the entire genome—a far more precise definition of the Arabidopsis centromeric regions than earlier estimates of more than 40% of the genome, says Preuss. Inside these regions as well as flanking them, Preuss and her team found more repetitive sequences, most notably, recognizable sequences of 180 base pairs repeated hundreds of times, on all five chromosomes. Surprising nuggets also turned up inside the centromeric regions, most notably, a significant number of genes. “That was one of the big interesting things,” says Preuss, because the centromere had long been considered relatively barren territory. The analysis has located some 200 genes in the Arabidopsis centromeric regions, at least 50 of which are expressed. About 40 of these genes appear only once in the genome sequence. “These are bona fide genes that would have been left out” without delving into the centromeric region, says Preuss. ## Creating chromosomes Having defined the centromeric regions and sequenced most of them, the researchers still need to find the functional part and figure out how it works. Given the diversity of centromeres in different organisms, there isn't a universal code to look for. To find the functional centromere, Preuss's lab is developing a new tool— experimental minichromosomes that are a stripped-down version of an Arabidopsis chromosome. They contain all the essential parts: the centromere; telomeric DNA from the chromosome ends; genes of interest or indicator genes, such as green fluorescent protein; and elements to ferry the package into cells. When assembled, these minichromosomes should function in plants alongside the other chromosomes. Postdoc Kevin C. Keith in Preuss's lab is now testing pieces of DNA from the Arabidopsis centromeric regions to see which ones work in cell division—the hallmark of a functional centromere. Keith is combing through all the bases in the centromeric region, including the 180-base repeats. The process is akin to going through jars of bolts in a hardware store to find the right one, in this case, a stretch of DNA that holds the minichromosome to the cell's protein motors. To do so, Keith is methodically inserting sequences of fewer than 100 kilobases into the minichromosomes to test which ones work as the centromere. Such minichromosomes are in the works for other organisms, including humans. Apart from their use as a tool to explore chromosomal functioning, they have an applied side in genetic engineering as well. Researchers believe they will provide a controlled means of “stacking” large numbers of genes—say, for pathogen resistance— into an organism that could also be engineered to be eliminated when necessary. One of Preuss's next interests is to use the minichromosomes to study why chromosomes don't always segregate properly and end up with two or no chromatids in a gamete instead of just one. If the gamete is used in fertilization, the resulting offspring will have an abnormal number of chromosomes, a condition known as aneuploidy. It happens between 1% and 2% of the time in Arabidopsis—about the same frequency as in yeast—but far more often in human meioses. The development of minichromosomes over the next 5 years “will open up tremendous resources” for exploring such phenomena, she says. Preuss isn't the only one who thinks so. This past summer she became a Howard Hughes Medical Institute investigator, only the third plant biologist to be so recognized. To commemorate the occasion, Preuss's lab presented her with a framed collage of the pollen tetrads and the data they've generated. At the center is a photo of Howard Hughes holding a fistful of Arabidopsis. It's in bloom. 17. CONSERVATION BIOLOGY # Galápagos Station Survives Latest Attack by Fishers 1. Dan Ferber Researchers at the Darwin Research Station put the pieces back together after a festering dispute over fishing quotas turns violent Botanist Alan Tye had a tough first week on the job after becoming acting director of the Charles Darwin Research Station in the Galápagos Islands in November. On Tuesday, he watched police put up barbed wire barricades after the research station on Santa Cruz was threatened with attack from local fishers. On Wednesday the fishers, angry over a quota on spiny lobsters that they feel is too low, hijacked Tye's dinghy during his commute to work. On Friday, Tye learned that Ecuadorian navy special forces had rescued two lab employees who, fearing for their lives, had taken refuge in mangrove swamps on Isabela, one of the station's three island sites. For Tye and others who have spent years tending to the famed tortoise population and performing other conservation studies, the week of 13 to 17 November was the latest reminder of their precarious existence on this research outpost in the eastern Pacific. In 1995, fishers armed with clubs and machetes took researchers and their families hostage after authorities stopped sea cucumber fishing for the year (Science, 3 February 1995, p. 611). In 1997, a park ranger was shot after wandering into an illegal fishing camp. And earlier this year, fishers angry about sea cucumber quotas took several endangered tortoises hostage. Yet the station has survived for 36 years. “We've been through this before,” says Tye, who hopes for peace on the 8000-square-kilometer archipelago, which lies 1000 kilometers west of the Ecuadorian mainland. “It's difficult at the time, but the experience throws us together and makes us more determined.” The fuse that set off the most recent conflagration was an annual 50-ton limit on spiny lobsters that local fishers reached barely halfway into the 4-month season. The quota had been set under a 1998 law that requires park authorities to consult with fishers, tourist service operators, and local officials and to draw on scientific advice from the Darwin station (Science, 20 March 1998, p. 1857). The law gives the National Park Service the authority to enforce the quotas, which are monitored by the research station, but when park officials moved in, the fishers reacted with what they term a “strike.” Unruly bands laid siege to the station and the park service, blocked roads and offices, tore down the island's telephone antenna, and destroyed research records. The park service bore the brunt of the fishers' anger. On the remote island of Isabela, where the two lab employees and park superintendent Juan Chavez were rescued from the swamps, rampaging fishers carried off computers and scrawled death threats on the walls. “They completely destroyed our office and burned absolutely everything,” says park spokesperson Desiree Cruz in an e-mail. They also threatened Chavez's life and trashed his home. In other areas, fishers blocked tourist boats from landing, and a local school official who sided with the fishers threatened high school students who had written letters supporting conservation efforts. “Some of this protest activity approaches terrorism,” says Darwin Station ecologist Howard Snell, who also teaches at the University of New Mexico, Albuquerque. Some research at the station was affected. During a 10-day occupation of station offices on Isabela, hair dryers that kept tortoise eggs warm enough for embryos to develop were taken. Many of the eggs “will possibly die,” including several embryos of critically endangered populations, according to recent e-mails to Snell from Ecuadorian herpetologist Cruz Marquez. The fishers also destroyed tortoise pedigree records, which ensure that the different island tortoise subspecies remain purebred. But because the breeding program hatches several hundred tortoises a year, Snell says, the damage was relatively minor. The 220 employees and volunteers at the research station are frustrated by recent events but have resumed their work, which includes efforts to save endangered native plants, control introduced plants, track the rich marine ecosystem, and breed endangered tortoises and iguana subspecies endemic to the islands, Tye reported in a 7 December e-mail. Research station staff and park workers also held a peaceful march on 23 November in Puerto Ayora, the largest town in the islands, to protest the violence and rally support for conservation. But the rally was postponed for 6 days because of threats of violence from local fishers. The biggest casualty of the riots could be the current conservation policy, which was widely seen as a way to include all interested parties. But enforcement against overfishing of lobsters, sea cucumbers, and long-lining for shark fins has been weak, often ignoring the limits. “Within the conservation community, there's a tremendous amount of frustration and disappointment,” Snell says. Meanwhile, an uneasy calm prevails. Park officials have petitioned Ecuador's president, Gustavo Noboa Bejarano, for protection against what they call “ecological terrorism,” and the fishers, while busy fishing, have threatened to renew their strike. The Ecuadorian government jailed three of the fishers, charging them with terrorism, and Noboa has promised that any convictions will be “sanctioned with the full force of the law.” Many people are skeptical, however. “We've had lots of words. What we need is action,” says a Western diplomatic source in Quito, who spoke on condition of anonymity. One environmental group, the Sea Shepherd Conservation Society, is taking action. Society president Paul Watson and his crew set sail for the Galápagos on 7 December from Los Angeles in Sirenian, a 29-meter former U.S. Coast Guard vessel. Carrying replacement computers and other supplies donated by U.S. scientists, the ship is scheduled to arrive this weekend. Park personnel, aided by three armed members of the Ecuadorian navy, will use it to patrol Galápagos waters for 5 years. The team plans to fight poaching by “confiscating” illegal fishing boats and their cargo. “This is a crisis situation,” says Watson. “If we can't save the Galápagos, what the hell can we save?” 18. BIOMATERIALS # New Chinese Biochip Center Straddles Business, Academe 1. Ding Yimin* 1. Ding Yimin is a reporter for China Features in Beijing. With reporting by Jeffrey Mervis. China taps a U.S.-trained entrepreneur and researcher for a new enterprise intended to lead the country into the big leagues of a burgeoning field BEIJING— Meet Cheng Jing, the new face of Chinese science. As a U.S.-trained researcher, head of an academic center-cum- commercial enterprise in Beijing, and founder of a biotech company in San Diego, the 37-year-old Cheng wears many hats. But that's exactly the mix of experience Chinese officials are looking for to help lead the country into one of the hottest areas of biotechnology. Last month, Cheng was named head of a new, two-pronged biochip venture here that will be highly unusual even for China, which has been experimenting with academic-industrial arrangements in recent years. One part is a for-profit company backed by$48 million from a combination of national and academic partners and overseas venture capital. The second piece is a nonprofit national center with nearly $10 million in research funding from the Ministry of Science and Technology. “Our goal [in creating the venture] is to catch up with the world's most advanced level of biochip development,” says Wang Li, an official from the Ministry of Science and Technology. And he says Cheng's U.S. experience was a big plus: “One of the measures we have taken is to support those scientists who came back from overseas with new ideas and skills.” The new biochip enterprise is tied to the top-rated Tsinghua University. Its two titles—the Capital Biochip Corp. and the National Center of Biochip Engineering—reflect its dual roles as research enterprise and commercial company. Cheng's mission is to train the next generation of biochip scientists and technicians and to develop new technologies that would be patented and licensed to other companies—including Aviva Biosciences Corp., the start-up that Cheng founded in San Diego. Bankrolled by nearly$30 million from the State Development Planning Commission and four academic partners, the corporation has also received $18 million in domestic and overseas venture capital. The science ministry is expected to kick in$9.6 million from a national high-tech program to fund research at the center.

Cheng will operate Capital Biochip along Western lines, including offering stock options to employees after 1 year on the job. He also hopes to hire 30 senior scientists and engineers from overseas by matching their current salaries. Within 3 years, he anticipates having 300 employees, half of them graduate students, ensconced in a 20,000-square-meter building that should be completed by early 2002. He expects most of his current staff of 30 to join the new national center, along with people from the other institutional partners—Huazhong University of Science and Technology, the Chinese Academy of Medical Sciences, and the Academy of Military Medical Sciences.

Cheng has plenty of experience in straddling the worlds of academe and commerce. He was trained as a railway engineer in Shanghai, got a Ph.D. in forensic science in the United Kingdom, and then moved into the biochip business at the University of Pennsylvania. “He was the kind of postdoc that you dream about—motivated, skilled, and someone who knows exactly what he wants,” recalls Penn's Larry Kricka, a professor of pathology and laboratory medicine. “He also understood the value of exploiting intellectual property. And that's key to what he's trying to do at the new center.”

Cheng left Penn for the San Diego-based Nanogen, where he helped develop a bioelectronic chip that isolates and purifies DNA and RNA from a whole blood sample. But after 3 years there he wanted to be his own boss and to operate on a larger scale. So in early 1999 he returned to China as a full professor at Tsinghua, which gave him a generous budget to set up an R&D center there. “He's had tremendous support that would make many of us in the U.S. jealous,” says Peter Wilding, a professor of clinical chemistry at Penn whose work over the past decade on capillary electrophoresis helped to lay the groundwork for “PCR-on-a-chip” technology.

Within months, Cheng had persuaded Tsinghua to back Aviva, which he set up down the road from his former employer. Its research staff has focused on technology that prepares the samples for analysis—the first and most difficult step in the process. In October the company unveiled its latest technology at an international biochip symposium in Beijing—a multiforce active biochip that uses micromagnets, acoustic, and dielectrophoretic forces to help isolate the material to be analyzed. In a clear demonstration of its significance, the meeting was a subset of an international conference opened by Chinese President Jiang Zemin. University officials have insisted that he spend half his time at Aviva, Cheng says, to make sure that it remains on track in licensing new technology and developing products.

Some scientists wonder if Cheng's new venture will be nimble enough to keep up with the latest technology, however. “It's not a good idea to develop the biochip industry by injecting a large amount of money in start-up funds,” says Lu Zuhong, head of a biochip laboratory at Southeast University in east China's Nanjing. The key to commercial success is not a large laboratory with lavish facilities, Lu says, but quickly converting a research finding into a marketable product.

Other scientists worry that Cheng may be spreading himself too thin. The national center has targeted half a dozen areas, ranging from refining the processes underlying any microlab on a chip to creating an implantable chip for therapeutic and monitoring purposes. But Hu Gengxi, a research professor at the Shanghai Institute of Cell Biology of the Chinese Academy of Sciences, thinks that it would be better to focus on one or two promising areas.

Cheng defends his approach, saying that one goal of the center is to “provide a more solid basis for development” of the Chinese biochip industry. Its successful dissemination of new technologies, he adds, “will benefit other, smaller Chinese companies” that might otherwise not have access to them. And Cheng says that a generous budget offers him the freedom to tackle several research questions simultaneously. Cheng also says that he'll try to minimize the potential conflicts of interest stemming from all the hats he is wearing. “Although I am chief technology officer for two companies, the ideas generated during my work and discussions in the United States will belong to the American company, while the national center will benefit from my ideas and service in Beijing.”

In tackling these and other issues, Cheng says there are few domestic models that he can follow. “We are making it up as we go along,” he confesses. But he wouldn't have it any other way. “We need to find the right approach that works for China,” he says. “You can't simply copy Western practices.”

19. DICK MOL

# 'Sir Mammoth' Leads Charge to Uncover Ice Age Fossils

1. Richard Stone

Dick Mol may be an amateur, but he's had more success than most professionals in his chosen field of paleontology

ON THE EASTERN SCHELDT, THE NETHERLANDS— In the driving rain on a recent autumn day, several slicker-clad men and women stand on the deck of the Dutch mussel cutter ZZ10. A clanging winch hauls up a dredge net and swings it over the side. Out spills a marine cornucopia: loads of brittlefish and mussels, a few flounder, the odd worm with iridescent bristles. The crew paws through the writhing mass, pushing aside the living creatures in search of something that died ages ago.

The Tiglian-type sediments in these waters off Zeeland date from 1.6 million to 1.8 million years ago, when today's estuary was dry land inhabited by southern mammoths, squat mastodons, giant deer, and the saber-toothed cats that preyed on them. Every year for the past half-century, a gang of scientists, amateur enthusiasts, and local officials has spent a day dredging for new fossils from this exotic menagerie to honor the fishers who spend their lives on these waters hauling up relics of a long-lost era. This time, one small bone fragment defies identification—at least until one of the sharpest eyes aboard the ZZ10 comes over to take a look. Stocky, blond Dirk Jan (Dick) Mol picks up the black bone, so heavily mineralized that it emits a sharp ping when tapped, and within a few seconds concludes that it's a fragment of a foot bone from a southern mammoth—not bad for a customs officer at Amsterdam airport. But not surprising: As paleontologist Jelle Reumer, director of the Natural History Museum in Rotterdam, explains, Mol is “Mr. Mammoth.”

No other country, perhaps, embraces amateur paleontology as warmly as the Netherlands does. “Vertebrate paleontology as an academic subject now hardly exists here,” says Reumer, who notes that hot fields such as genetics tend to get the few new academic positions created at universities in his country. “Amateurs help fill that gap.” Mol may be the most accomplished of the amateurs, says John De Vos, a curator at the National Museum of Natural History in Leiden: “He knows every mammoth specimen in Europe. He's crazy! He's obsessed!”

The 45-year-old Mol is also a celebrity. As scientific coordinator of a major expedition that's gathering the remains of woolly mammoths and other Pleistocene fauna from Siberia's Taimyr Peninsula, Mol has been featured in a documentary on the Discovery Channel and in a sequel to appear next March. The work has brought him international recognition for his studies on quaternary paleontology, the study of the Pleistocene and today's Holocene Epochs. He's even earned a knighthood from Queen Beatrix.

Mol's fascination with fossils was kindled when a grade school geography teacher in his hometown of Winterswijk showed him a collection of sea urchin fossils dating from tens of millions of years ago. As a teenager, he went with his uncle to the Leiden Museum, where curators showed him backroom collections of mammoth bones gathered from the North Sea. “I decided from that moment on to collect the remains of ice age mammals,” he says.

The eighth of nine children, Mol could not afford to attend university. But his decision to join the customs service in 1974 proved to be a major boon to his avocation. That same year, the Netherlands implemented the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES), which gave customs officers sweeping interdiction powers. Mol, who trained as a CITES specialist, spends so much time on the job studying bones and interacting with pros that he doesn't regret missing out on an academic career. Nor has he ever found a colleague from academia reluctant to collaborate with him. “Many scientists don't even know that I'm a customs officer,” he says.

His employer has also benefited from his passion. A traveler arriving from Hong Kong once declared that an ornate sculpture of a Chinese god was carved from woolly mammoth ivory. Because mammoths are extinct, their parts are not protected under CITES, unlike those of living elephants. “I told him, ‘I do not believe you, and I'll tell you why,’” Mol recalls. The object was clearly carved from a straight tusk without any cracks, he says—a dead giveaway, considering that mammoth tusks are curved and even the best preserved tusks from the Siberian permafrost have cracks. Mol gave the chagrined man a choice: Either ask a court for a radiocarbon date or pay a steep fine. The would-be smuggler paid up.

As Mol's self-taught acumen grew, he started to see things that professionals had missed. He learned from fragments snared in fishing nets that the North Sea faunal assembly features three mammoth species, illustrating that the sediments encompassed millions more years than previously thought. His work has generated dozens of scientific papers and collaborations with top academic researchers. He was also the first to find a flaw in a highly publicized 1993 report in Nature claiming that the freshest mammoth remains ever found—bones of individuals that died only 3800 years ago on Siberia's remote Wrangel Island—belonged to dwarf male mammoths. After seeing the bones, Mol realized they were not dwarfs but old female mammoths, which are smaller than their male counterparts; that left bragging rights to the only true dwarf mammoths known from the fossil record to the Channel Islands off California.

Over the years, Mol has amassed about 15,000 specimens in his suburban townhouse. His most beautiful trophies—including exquisitely preserved mammoth tusks, teeth, and jawbones and a Pleistocene musk ox horn sheath—litter the bookshelves and floors. The rest, cleaned, ID'd, and numbered, fill some 600 footlocker-sized Styrofoam boxes. When Mol's daughter moved to The Hague a few years ago, her room was soon taken over by bones. “I'm waiting for my son to leave as well,” he says with a laugh.

Mol's big break came in August 1998 with a call from Bernard Buigues, a Frenchman who runs a Paris-based tour company specializing in trips to the North Pole. Earlier that year, Buigues said, he had discovered the remains of a woolly mammoth at a site on the Taimyr Peninsula, far above the Arctic Circle. Buigues wanted Mol's help to excavate the mammoth. The result was a remarkable collaboration involving two dozen scientists from around the world.

Buigues had planned a traditional excavation that would blast the permafrost with hot water in summertime. “But that would have destroyed the scientific information,” says Mol, who persuaded the Frenchman to undertake a riskier endeavor: chiseling out the chunk of permafrost containing the remains and airlifting the block to a cold room for study (Science, 29 October 1999, p. 876). The mammoth is now being thawed slowly so that any flesh in the block can be recovered still frozen and associated plants, pollen, and insects can be retrieved. “It will be the first time that scientists have access to materials in which the frozen chain has not been broken,” Mol says.

Last summer, Buigues, funded royally by the Discovery Channel, mounted the largest expedition ever to collect mammoth remains in Siberia. The researchers hope to shed light on the shifting mosaic of species, from mammoths and woolly rhinos to saiga and voles, that lived throughout the late Pleistocene and early Holocene. They also hope to test a provocative theory that mammoths and many other large animals succumbed not to climate change or human hunters—as most researchers think—but rather to an apocalyptic plague.

Particularly noteworthy is the expedition's leadership. While the chief scientist is renowned paleoanthropologist Yves Coppens, who co-discovered the Lucy hominid bones, the key decisions are made by two amateurs. Buigues oversees logistics, and Mol sets the scientific agenda with pro forma approval from Coppens, who has visited the site but has not participated in the fieldwork. Mol's stewardship of the program—and his avowed skepticism toward the possibility of cloning the mammoth —impresses other expedition members. “Dick is a very, very smart guy. But because he never got a position in academia, he has to wear this big ‘A’ for amateur on his forehead,” says mammalogist Ross MacPhee. A curator at the American Museum of Natural History in New York City, MacPhee and Preston Marx of the Aaron Diamond AIDS Research Center in New York City proposed the plague theory 3 years ago. MacPhee is impressed with the openness with which Mol has conducted the program, including a plan to distribute mammoth remains to anybody who submits a serious research proposal.

Mol's most lasting legacy may be the credibility he lends to a thriving community of amateur paleontologists. Last May, when Mr. Mammoth became Sir Mammoth, the honor brought practical results. “I asked my boss if it would be possible to get some extra holiday to spend in Siberia with the expedition,” says Mol, who had already used up his generous allotment of 3 months annual paid leave. “He said, ‘Of course. How much time do you need?’”