News this Week

Science  23 May 2003:
Vol. 300, Issue 5623, pp. 1212
1. AIDS RESEARCH

Drug Trials Without the Drugs?

1. Jon Cohen

Luis Montaner, an AIDS researcher at the Wistar Institute in Philadelphia, Pennsylvania, received word last month from the National Institutes of Health (NIH) that a grant proposal he had submitted to stage a novel drug study in South Africa had come through peer review with flying colors. But rather than celebrating their success, Montaner and his collaborators find themselves all dressed up with nowhere to go: They can't afford the fare. They have been told that NIH will not pay for the anti-HIV drugs the researchers are planning to test.

Montaner's project, like several others in the pipeline, is caught in a kind of Catch-22. NIH's main AIDS branch, the National Institute of Allergy and Infectious Diseases (NIAID), wants to encourage research on the best ways to deliver AIDS therapies in developing countries, but it is concerned that the cost of buying drugs for such studies would swamp its research budget. This concern is spelled out in a draft document now circulating at NIAID, a copy of which has been obtained by Science. The document is sparking a spirited debate among AIDS researchers.

The draft policy states that the cost of purchasing drugs for clinical trials “would severely restrict the Institute's research capacity by limiting the number, scope, duration and focus of NIAID's international HIV-related research activities.” It adds that only “under extraordinary circumstances” would NIAID foot the bill for test drugs. The document goes on to argue that it would be unethical to stop treatment when a trial ends. It says researchers should therefore be required to submit a plan “to provide appropriate care and [antiretroviral] treatment for study participants after completion of the trial.” None of these requirements are new, the document notes: They simply formalize NIAID's existing policies.

Montaner's study crystallizes many of these issues. He is hoping to test whether there would be any downside to occasionally stopping anti-HIV drugs in a structured treatment program. If a stop-start regimen works as well as continuous treatment, patients could take one-third fewer drugs—which could amount to a huge cost savings in poor countries.

He says reviewers of his proposal and NIH staff members encouraged him to remove the cost of test drugs from earlier versions of his grant request. That means he would either have to find another source of funds to pay for the drugs or persuade companies to donate them, both of which he has just started to do. That wouldn't be a problem for clinical trials that are required for regulatory approval; companies are eager to donate drugs for such tests. But Montaner is planning to test regimens involving already approved drugs—and his study could result in cheaper treatments, which could cut into company profits.

Montaner argues that “the drug costs should be addressed as any other research costs, as long as they're justified from ethical and research perspectives.” Ed Tramont, head of NIAID's Division of AIDS, says NIAID will help investigators like Montaner find drugs, and he is exploring the possibility of a foundation acting as an intermediary with the drug companies. He agrees that “the incentive is not there” for companies themselves to donate drugs for such trials. And the requirement for post-trial treatment is an added disincentive, he says: “It's fear of an open-ended, longtime commitment.”

Montaner isn't the only researcher facing such problems. An NIAID-funded drug testing network called the AIDS Clinical Trials Group (ACTG), in conjunction with the HIV Prevention Trials Network (HPTN), hopes to test the therapeutic effects of various combinations of drugs and assess whether patients become less infectious. Four companies have agreed to supply drugs for the study, which is scheduled to take place in 12 different countries. But the researchers still need two others, as well as backup drugs to help people who become resistant to the first-line treatments. “We're looking at donor agencies to cover the gap,” explains the principal investigator of the HPTN study, Myron Cohen of the University of North Carolina, Chapel Hill. The cost could be substantial. Researchers are reluctant to use steeply discounted generic versions of the drugs because they want to avoid questions about potency and other manufacturing issues.

Both ACTG and HPTN are trying to figure out how to treat people after the studies end; neither has solved the puzzle yet. “Everybody wants to do the right thing,” says Cohen. “We're trying to be good citizens, but it creates the situation where people may never get drugs.”

AIDS immunologist Bruce Walker of Massachusetts General Hospital in Boston is facing similar difficulties. “Right now, there are plenty of groups like ours that are ready to treat people, and we can't get drugs,” says Walker, whose study in South Africa has been on hold for a year. He is particularly worried that far-removed ethics concerns may affect “operational” research aimed at answering practical—and often unexciting—questions about how best to treat people in different settings. “The absurdity of the situation is that 95% of HIV infections exist in countries where you have minimal experience giving the drugs. There's a real problem here,” he says.

Walker calls the NIAID draft policy “not very visionary,” and he predicts that it will “stifle progress.” And he takes strong exception to the suggestion that researchers must figure out how to provide treatment after a study ends. “We're letting a lot of people die because we're saying [you must treat] forever,” Walker says. “We have plenty of people who were dying who are now alive because they're on therapy. People would rather be alive and faced with having to figure out what they're going to do in 3 years than be dead.”

Tramont says NIH's Office of AIDS Research is now reviewing the policy. “It's very complicated,” he concludes, “and has a lot of implications.”

2. EUROPEAN SPACE AGENCY

Financial Crisis Puts Comet Mission on the Ropes

1. Daniel Clery*
1. With reporting by Giselle Weiss in Allschwil, Switzerland, and Govert Schilling in Utrecht, Netherlands.

CAMBRIDGE, U.K.—For scientists on the Rosetta mission, chasing around the solar system after a comet and dropping a lander onto its surface will seem like a cakewalk after the trouble they've had getting off Earth's surface. After the failure of an Ariane 5 rocket last December postponed its launch and nixed its target comet, Rosetta appeared to be back on track last week when the European Space Agency's (ESA's) Science Programme Committee formally settled on a new target, comet Churyumov-Gerasimenko (Science, 14 March, p. 1638). But the revised launch date of February 2004 is far from certain. As Science went to press, a combination of crises had landed the science program $230 million in the red this year—an amount equal to more than half its$430 million budget. The only expendable item is Rosetta. “This is an unexpected final hurdle for Rosetta to cross,” says David Southwood, ESA's science chief. And the only way to surmount it, he says, will be if ESA's governors agree to break their own rules.

The $1 billion Rosetta was originally meant to set off last January for a rendezvous with comet Wirtanen in 2011. The latest plan, to launch in February, would allow Rosetta to catch up with its new target, Churyumov-Gerasimenko, in late 2014. “Everything is technically possible,” says David Hall, director of the British National Space Centre and a Science Programme Committee member, although Rosetta “may have a difficult landing.” Churyumov-Gerasimenko is bigger than Wirtanen, so the lander is being modified to cope with stronger gravity. But Rosetta may come down with an even bigger bump if ESA's science program can't find a way out of its budget problems. To squeeze the maximum science out of a limited budget, the program does not have much money lying around for contingencies. Since Rosetta's grounding, it has been “kept alive with a month-to-month drip feed” of funding, says Southwood. Altogether, the cost of storing Rosetta and launching next winter, along with the impact of the delay on other missions—including Mars Express, Venus Express, and SMART-1—is about$90 million.

That's not the only bill the science program is having a hard time paying. Europe's struggling aerospace industry has been pressuring ESA for early payment on some contracts, to the tune of an extra $60 million. And ESA has been forced to bail out national space agencies that normally pay for the instruments on ESA missions. The science program has forked out about$80 million on instruments, mostly for the Herschel and Planck observatories. “The science program could reasonably deal with any one of these problems,” Southwood says. But juggling three debts is proving well-nigh impossible. Most of this year's science budget has already been paid out to contractors for future missions; the only sizable chunk of money uncommitted is that needed to get Rosetta into space. Nevertheless, the Science Programme Committee last week renewed its commitment to Rosetta. “The ball is now in my court to find a budget,” says Southwood.

“Within the current ESA rules, there's no easy solution,” says Gerard Schwehm, Rosetta's project scientist. The main problem is that ESA's rules prohibit it from taking out bank loans to cover overruns or shifting money between program budgets, which are set by government ministers. According to Hall, the science program will ask to borrow funds from other ESA directorates to cover the last 2 months of 2003. “Some programs do underspend,” says Hall. But this kind of maneuvering, he adds, has never been done before at ESA, “and some may not be keen to set a precedent.”

Southwood was due to present the borrowing plan earlier this week to ESA's administration and finance committee, which would decide on its feasibility. If approved, the proposal would go before the ESA Council next week or in June. Southwood is eager for a quick resolution: He says he has to take action next month to ensure that Rosetta's prime contractor can get everything ready for a February launch. However, Hall cautions, “it is not a foregone conclusion” that national representatives on the council will support the rescue plan.

3. CLIMATE

A Surprise La Niña

1. Richard A. Kerr

Government climate experts announced this week that El Niño's opposite number, La Niña, seems to be emerging in the tropical Pacific Ocean. La Niña's unusually cool waters signal weather extremes around the world, although these are less troublesome than those brought on by El Niño's warmth. Perhaps the hardest hit are the forecasters: Neither they nor their computers saw La Niña coming. “We can't get away from the fact that we didn't predict it,” says meteorologist Anthony Barnston of Columbia University's International Research Institute for Climate Prediction (IRI) in Palisades, New York. “Collectively, these models aren't handling the situation as well as humans can,” and this time humans seem to have blown the forecast too.

Until recent weeks, the Pacific had seemed to be behaving predictably enough. Human forecasters, if not the computers, had anticipated last winter's El Niño the previous winter—half a year before its onset (Science, 26 July 2002, p. 497). Humans had even gotten its eventual moderate magnitude right. And through this March and into April, El Niño had been fading away pretty much by the book. In March, meteorologist Vernon Kousky and his colleagues at the National Weather Service's Climate Prediction Center (CPC) in Camp Springs, Maryland, did note that some ongoing cooling in the eastern tropical Pacific “supports the possibility of the development of La Niña later this year.”

Despite that early hint, by mid-April both CPC and Barnston and his colleagues at IRI went with much the same forecast: Near-normal conditions through September were likely. Sixteen different forecast models that were run at institutions around the world in late April and early May didn't help much; six called for neutral conditions the next 6 months, four predicted another El Niño, and five the next La Niña.

Now, however, rapid developments in the tropical Pacific have driven forecasters to revise their recent forecasts drastically. The cooling glimpsed in March has taken off, chilling the upper few hundred meters of the eastern and central Pacific. On 16 May, IRI upped its chances of a full-blown La Niña this fall from 25%—essentially a random guess—to 55%. On 19 May, Kousky said that “oceanic and atmospheric conditions indicate that a transition to La Niña is already under way.” The CPC forecast now stands at a 70% chance.

If the tropical Pacific follows through as predicted, it will mean a mixed bag of weather. La Niña would encourage the development of hurricanes in the tropical Atlantic this summer and fall, meteorologist Gerald Bell of CPC said today in conjunction with a National Oceanic and Atmospheric Administration forecast of above-normal hurricane activity. But other parts of the world could benefit from La Niña. India would be less likely to suffer a failure of its crucial summer monsoon, Barnston notes. Next winter, Alaska, which has sweltered under unusually mild winters in recent years, could get back to normal, says Kousky. Last winter, during the El Niño, the renowned Iditarod dogsled race there had to be relocated northward for lack of snow.

The failure to give any more warning this time around seems to lie with the nature of the El Niño-La Niña cycle. Especially in the spring, when the tropical Pacific is most likely to switch from one mode to the other, “nuances can have large effects on the outcome. The butterfly effect is still there,” says Barnston, referring to the analogy often used to describe chaotic systems: A butterfly flapping its wings in Brazil sets off a tornado in Oklahoma. The tropical Pacific monitoring system is letting researchers consistently pick up the earliest signs of a switch, but some transitions are still tough to forecast, even for humans, who are still able to weigh the evidence of impending shifts more subtly than the machines can. “We've come a long way in 20 years,” says Kousky, “but we've got a ways to go.” Another 10 years, he says, may pass before reliable forecasts are more the norm.

4. NEUROSCIENCE

Monarchs Check Clock to Chart Migration Route

1. Elizabeth Pennisi

People who need to consult maps, radio traffic reports, or the Global Positioning System to navigate from one city to another should stand in awe of monarch butterflies. They migrate thousands of kilometers to a small winter retreat in Mexico. These intrepid travelers use the sun as a guide, but exactly how has been a mystery. Now researchers demonstrate that monarchs depend on an internal clock to determine their course.

Studies of birds 50 years ago indicated that long-distance migrants “need an accurate internal timepiece” to use the sun as a compass, says Steven Reppert, a neurobiologist at the University of Massachusetts Medical School in Worcester. Without a clock, the sun would prove an unreliable landmark as it moves across the horizon; a clock allows animals to compensate for this apparent motion and maintain a direct course.

By combining genetic and behavioral studies, as reported on page 1303, Reppert and his colleagues established that monarchs use their internal timepiece, apparently set by genes that are important to circadian rhythms, to fly in the right direction. The work “shows provocative links between the circadian machinery and the compass mechanism required for seasonal migration,” says Gene Robinson, a neurobiologist at the University of Illinois, Urbana-Champaign.

To demonstrate that monarchs have internal clocks and that these clocks are set by daylight, Reppert's collaborator Oren Froy tested monarchs' emergence from their chrysalises. They normally emerge in the morning, but exposure to constant light instead of daily light-dark regimens caused the butterflies to emerge throughout the day, indicating that their clocks had been disrupted.

Next, Anthony Gotter, also part of Reppert's team, looked at flight patterns during lab simulations of a normal day-night cycle, one in which light was advanced 6 hours, and one in which night never came. Once the butterflies were accustomed to these light regimens, Gotter tested their behavior outside using a flight simulator, a 150-liter barrel in which the butterfly flies while tethered to a tungsten wire connected to a computer. The monarchs could track the sun's path as it moved over the top of the simulator, which was open to the sky. Butterflies whose lab light cycle corresponded to local daytime headed southwest. They had been collected in Massachusetts at the beginning of the study, and “that seems to be the genetically programmed direction that they need to go to get to Mexico,” Reppert says.

Next the researchers looked at monarchs whose exposure to light was advanced by 6 hours, so the “day” in the lab lasted from 1 a.m. to 1 p.m. Outside, these would-be migrants shifted their orientation by 115° and headed southeast. Like birds whose circadian rhythms have been fooled, the time-shifted insects failed to interpret the sun's position correctly. During the morning, they acted as though the sun were in the west rather than the east. Butterflies that had been in constant light indoors simply headed straight for the sun no matter where it was in the sky.

To see whether gene expression corresponded to light-induced changes in behavior, Reppert and his colleagues measured RNA from a gene called period (per). Fruit fly researchers had identified per as a key gene for setting the internal clock, and Reppert's group isolated the gene in monarchs. The per gene's activity was high during dark periods and low during light times. Butterflies exposed to constant light for several days showed no swings in per activity.

Taken together, the experiments demonstrate that the internal clock, perhaps timed by per activity, tells monarchs how to calibrate their movements against the sun. “It's the first molecular entrée” into the clock-assisted migration, says Charalambos Kyriacou, a behavioral geneticist at the University of Leicester, U.K.

But it's no slam dunk. Orley Taylor, an entomologist at the University of Kansas, Lawrence, is not sure the work demonstrates that the clock guides navigation. “The data would be more conservatively interpreted as phototactic [light-based] orientation,” he says.

In any case, many mysteries of monarch migration remain. For instance, monarchs from different parts of the eastern United States and Canada all end up in the same Mexican wintering grounds. They follow different bearings and start their journeys at different times, a feat that is hard to reconcile with a simple, species-wide clock, says Taylor. A further complication is that each year at least three generations of butterflies emerge before the last one flutters to Mexico in the fall. No one knows what triggers migration in that generation, says Reppert, but the circadian clock might help them recognize shortened day lengths and therefore the changing of the seasons.

5. SARS

Researcher Told to Stay Home After China Trip

1. Martin Enserink

NEW YORK CITY—Ian Lipkin is working hard to improve diagnostic tests for severe acute respiratory syndrome (SARS). And now he knows firsthand how important they are: An accurate test might have prevented health officials from slapping Lipkin, a researcher at Columbia University in New York City, with an isolation order last week, when he developed fever and a cough after returning from China.

Lipkin flew back to New York City on 7 May after spending 3 days with top SARS researchers in Beijing at the invitation of the Chinese government. Six days later, he developed symptoms that included fever and a cough, making him a suspected SARS case.

Lipkin is convinced he doesn't have the disease; his first symptoms were a runny nose, sneezing, and congestion, whereas SARS usually starts with a fever and an unproductive cough. What's more, four nasal and two oral swabs tested at his lab using the polymerase chain reaction all came back “stone-cold negative,” Lipkin says, “and I think I'm running the best assay on the planet.”

Nonetheless, Lipkin decided to stay away from a SARS meeting at the New York Academy of Sciences last Saturday that he had helped organize; instead, he followed the talks and asked questions through a speakerphone, occasionally startling participants with frightful coughs. Also at the meeting were Marcelle Layton and Donald Weiss, two top communicable-disease officials at the city's Department of Health. Before the day was over, Lipkin received a call from the department, telling him to stay home until 25 May.

New York City Department of Health spokesperson Andrew Tucker says the agency doesn't discuss individual cases, but he confirmed that a “50-year-old male with a travel history” became New York City's 22nd suspected SARS case this weekend. Allan Rosenfield, dean of Columbia's Mailman School of Public Health, says “everybody here is 99% sure [Lipkin] just had a bad cold.” But he says he respects the city's decision.

Lipkin, who had to cancel a talk at the American Society for Microbiology's general meeting on Tuesday, says his temporary imprisonment drives home the need to develop a validated test as soon as possible—the next flu and cold season could be “absolutely crippling to the health system” without one, he says. In the meantime, he's trying to work from home; and in case he gets bored, he says, “Marci Layton has said she'll rent me some videos.”

6. ROYAL SOCIETY

Nine Women Make 2003 a Record Year

1. Yudhijit Bhattacharjee

A record nine women are among the 42 new fellows elected by the U.K.'s Royal Society this year. Monday's bumper crop of female fellows comes just 3 weeks after the society's U.S. counterpart, the National Academy of Sciences (NAS), elected its highest-ever number of women—17—in its annual class of 72 fellows (Science, 9 May, p. 883). Women now make up 4.4% of the Royal Society's total fellowship of 1290, and 7.7% of NAS's 2015 members are female.

“The underrepresentation of women in science, engineering, and technology remains a major problem, but progress is being made,” said Robert May, president of the society, in a press statement (http://www.royalsociety.org/). He said that 11% of fellows elected in the past 5 years have been women, “which runs somewhat ahead of the percentage of women professors” in scientific disciplines at U.K. universities.

The 343-year-old scientific academy, which has never before elected more than four women in a year, came under fire from a British parliamentary committee in 2002 for having a low proportion of women and ethnic minorities. It has defended itself against charges of gender bias by pointing to the lack of women in U.K.'s scientific workforce. Julia Higgins, who was elected in 1995 and is now a vice president, says the society cannot be expected to elect a large share of women when “the pool of women at senior levels in science in the U.K. is so small.”

Among this year's fellows is Jocelyn Bell Burnell, a University of Bath astronomer whose doctoral work at Cambridge was a cornerstone in the discovery of pulsars in 1968. The discovery won a Nobel for her adviser, Antony Hewish, who shared the prize in 1974 with another British astronomer, Martin Ryle. A number of prominent scientists maintain that Burnell deserved the prize as much as Hewish and Ryle.

“The perception back then was that science was done by senior men who had a fleet of minions to do their bosses' bidding,” says Burnell. “When the Nobel was awarded, I wasn't in the frame because I was a junior and a woman.” The increase in the percentage of women fellows, says Burnell, reflects a broader trend of many women scientists, such as DNA crystallographer Rosalind Franklin, “getting written back into the history of their subject.”

Other women joining the fellowship include Oxford molecular biologist Kay Davies, Harvard developmental geneticist Elizabeth Robertson, and Cambridge mechanical engineer Ann Dowling. Among the society's newly elected foreign members are two men from Germany, two from the United States, and one each from Mexico and Switzerland.

7. SOLAR CELLS

Tricks for Beating the Heat Help Panels See the Light

1. Robert F. Service

Solar power visionaries have long dreamed of solar panels that are cheap, efficient, and able to withstand the relentless heat of the midsummer sun. Twelve years ago, chemist Michael Grätzel of the Swiss Federal Institute of Technology in Lausanne and colleagues ticked off two of the three requirements when they introduced “dye-sensitized” solar cells, which use cheaper starting materials than the traditional silicon cells to generate power at moderate efficiency. Grätzel's solar cells, however, couldn't stand the heat. Now he's taken another step, adding substance to the dream of ultracheap solar cells.

In the 18 May issue of Nature Materials, Grätzel and colleagues in Switzerland and at Hitachi Maxell in Osaka, Japan, report success with a material that can withstand blistering temperatures for over a month at a time, a key industry benchmark. They did it by outfitting dye-sensitized solar cells with water-repelling, light-absorbing dyes and by using a novel charge-conducting electrolyte that is stable at high temperatures. “It's an important step,” says Brian O'Regan, an expert on dye-sensitized solar cells at the Energy Research Centre of the Netherlands in Petten. “This was one of the things blocking a lot of people from taking dye-sensitized solar cells seriously.”

The solar cell industry has grown by more than 20% a year for the past decade and last year shipped panels worth $3 billion total. But because of its relatively high cost, solar power accounts for only 1% of the U.S. electrical output. If the improved cells can be manufactured cheaply—still a big unknown—O'Regan says they could soon give traditional silicon cells a run for their money and make the industry more competitive with rival sources of power. All solar cells harness energy from the sun by exciting electrons in a semiconductor and shuttling them into an external circuit. In conventional solar cells, a semiconducting material—usually crystalline silicon—absorbs photons of light. The photons give electrons in the material an energy kick that makes them flow through the semiconductor to an electrode, where they can be connected to an electronic circuit and used to perform work. As the electrons race through the semiconductor, however, they risk running into “holes” left by other electrons. When that happens, electrons and holes can easily recombine, causing the electrons to shed their extra energy as heat. To give the moving electrons the clearest possible path to the electrode, the semiconductors in the best conventional solar cells must be highly pure and defect free—and thus expensive to produce. Dye-sensitized solar cells adopt a low-cost solution to the problem of purity by separating the two jobs of collecting light and ferrying electrons. At their core is a continuous network of titanium dioxide (TiO2) particles, each just a couple of tens of nanometers across, coated with a film of organic dye. The dye molecules absorb light and then quickly pass energized electrons to a neighboring TiO2 particle, where the electrons race through the particle network to the electrode. The holes are left behind on the dye layer, where they have little chance of recombining with energized electrons. Those holes are then filled with low-energy electrons that flow through an electrolyte from a counter electrode, allowing the cycle to be repeated over and over. The best dye-sensitized solar cells can convert more than 10% of the energy in photons into electricity, an impressive rate. But when the cells warm up, things can quickly go wrong. At high temperatures, dye molecules can fall off the TiO2 particles and dissolve into tiny pockets of water that has seeped into the cell. Conventional electrolytes break down and release carbon dioxide gas, which can build up and burst the cells apart. “It became a real problem to master the high-temperature stability problem,” Grätzel says. Grätzel's team hoped to solve both of these problems simultaneously. First, the researchers tagged each molecule with a pair of long hydrocarbon chains. The oily appendages made the dyes extremely hydrophobic and therefore less likely to give up their link to a TiO2 particle and dissolve in water. And for the electrolyte, Grätzel and colleagues turned to an “ionic liquid” that is stable up to 170°C. They also spiked the ionic liquid electrolyte with a gelling agent, creating a rubbery material that is far easier to work with during manufacturing. Together, the changes helped the cells reach an efficiency of 6% and endure a temperature of 80°C for 1000 hours while losing only 4% of their performance. That's well below the industry standard of 10%. The new cells do have a flaw, Grätzel acknowledges: Their current and voltage fluctuated during the high-temperature test, a characteristic that would make it hard to link cells together to make a solar panel. Grätzel says his team has since tamed the fluctuations, although he won't say how until his group can publish the results. Even without knowing the latest trick, companies are already gearing up to incorporate the modifications into products. Earlier this year Sustainable Technologies International of Queanbeyan, Australia, became the first company to market dye-sensitized solar cells. Managing director Gavin Tulloch says the company hopes to incorporate the modified dyes and gel electrolytes into its products. Another would-be manufacturer, Konarka Technologies of Lowell, Massachusetts, is adding the improvements to its prototype dye-sensitized solar cells, which it produces on a cheap, flexible polymer substrate. If the technology proves robust and reliable, it could produce solar cells at just one-fifth the cost of conventional silicon panels, says Konarka's research chief, Russell Gaudiana. That would still leave solar power more expensive than a large coal plant as a source of electricity, but it would give solar energy companies another major boost. 8. KAZAKHSTAN Plutonium Fields Forever 1. Richard Stone A decade after inheriting the Soviet Union's vast nuclear testing range in Central Asia, authorities in Kazakhstan are only now discovering the extent of a dangerous legacy underfoot: plutonium hot spots that pose a serious proliferation threat KURCHATOV, KAZAKHSTAN—At an undisclosed location in northeastern Kazakhstan, workers clad in white suits and respiratory masks are paving an area the size of a football field with a 2-meter-thick slab of steel-reinforced concrete. This covert operation, in the middle of desolate, shrubby steppe that once formed part of the Soviet Union's Semipalatinsk Test Site, might never have been contemplated before the 11 September terror attacks. But it has taken on a new sense of urgency. The reason: The soil is laced with plutonium. If inhaled, the plutonium dust, a Cold War leftover, would pose a cancer risk. But that's not the concern that is driving “Operation Groundhog,” which Kazakh officials described to Science. The nightmare is that terrorists could cart away the fissile material and fashion it into a “dirty bomb,” a device to disperse the plutonium using conventional explosives. Sophisticated atom thieves could even extract enough plutonium to construct “a small nuclear device,” asserts Timur Zhantikin, chair of the Atomic Energy Committee, Kazakhstan's nuclear regulatory body. Adding to the fears, says Kairat Kadyrzhanov, director of the Institute of Nuclear Physics in Almaty, is that “today you have essentially uncontrolled and unrestricted access to the test site.” It is no surprise that parts of the site—a territory nearly as big as Israel that once served as the Soviet Union's main testing grounds for atomic weapons—are badly contaminated. Over 4 decades, 456 nuclear detonations and dozens more “model experiments” were carried out here. What's disturbing is that until a few months ago, officials at Kazakhstan's National Nuclear Center (NNC), which manages the test site, knew nothing of the plutonium patch that they are now putting beyond use. Russian specialists divulged the information about the hot spot—located outside the four main testing areas—during recent trilateral talks between Kazakhstan, Russia, and the United States aimed at identifying proliferation threats at the site. The Russians are permitted to withhold classified information about the nature of the experiments or tests at Semipalatinsk. But Kazakh officials believe that the hot spot is the legacy of a model experiment in which physicists detonated a nuclear device that released little nuclear energy. And although test site managers have compiled crude maps of known plutonium contamination across the site, they expect that more such patches will come to light. “There are secret or hidden places where experiments were done,” says Gabriele Voigt, director of the International Atomic Energy Agency's Seibersdorf Laboratories in Austria, who has led radioecological surveys at the site. “I'm quite sure there are more to be found.” Authorities are taking no chances. The plutonium is “a serious proliferation concern,” says an official at the U.S. Department of Defense, which is footing the bill for Operation Groundhog. “Our most pressing problem is preventing terrorist access to fissile materials,” adds NNC Director-General Shamil Tukhvatulin. The disturbing picture of plutonium contamination that's emerging is a setback for Kazakh officials, who had hoped to transfer major tracts of the test site to local municipalities for farming, grazing, and mining during the site's decommissioning after the Soviet breakup. “That's what we thought 10 years ago, anyway,” says Tukhvatulin. Now he has a new mandate: keeping out interlopers while his teams root out hidden hot spots. “It's not like the material is lying there on the ground for us to find,” says Zhenis Zhotabaev, NNC's deputy director-general. A comprehensive survey of the territory is necessary, experts concur, but this could take years: Although contamination in the main testing areas is well documented, overall less than a third of the site has been scoured for radionuclides. Joining the club On an autumn afternoon in 1954, Zhotabaev was leaving the schoolhouse with his fifth-grade classmates in Chersk, a village north of Semipalatinsk, when a thunderous shock wave tore through, shattering the school's windows. “We didn't understand what had happened,” Zhotabaev says. The children knew it wasn't an earthquake; their favorite theory, he says, was that a new kind of Soviet jet had broken the sound barrier. Apart from a handful of top Communist Party officials and the villagers who lived on the test site's fringes, no one in Kazakhstan had a clue that atomic tests were being carried out about 150 kilometers southwest of Semipalatinsk, a city of 700,000. Even after joining the Institute of Nuclear Physics in the mid-1970s, Zhotabaev says he was privy only to the fact that the military operated a nuclear reactor near Semipalatinsk. Its purpose, like the rest of the site's activities, was classified. And questions were forbidden. It wasn't until perestroika in the mid-1980s that details about Semipalatinsk's secret history began to trickle out. Facility 905, as the test site was called when construction began in autumn 1947, was where Igor Kurchatov—the father of the Soviet bomb—and his comrades among the country's nuclear elite proved their mettle. More than 1000 top physicists and engineers ended up working in their purpose-built city of Kurchatov, curiously code-named Moscow-400 in the early days. When assigned here, nuclear talent fresh out of university would be elated to hear they'd be stationed near Moscow. “You can imagine their surprise,” recounts Aleksandr Kolbaenkov, chief engineer of NNC's Baikal-1 nuclear reactor complex, when the recruits, after arriving at the train station, found out that they were embarking on a 4-day journey to Central Asia. The first “physics package,” a plutonium bomb nicknamed RDS-1, for “Russia Does It Herself,” was detonated atop a 38-meter-tall tower at Ground Zero, the center of Technical Area Sh, on 29 August 1949. To probe the effects of the 22-kiloton blast, full-scale replicas of apartment buildings, airplanes, bridges, and tanks were arrayed at various distances from the tower. Some 1538 animals, including rabbits, sheep, and dogs, were caged in the test area; a quarter were incinerated, the rest observed for ill effects. The Soviet Union had become the second member of the nuclear weapons club, and it wasn't long before the two superpowers embraced the chilling doctrine of “mutual assured destruction.” Over the next decade, the Soviet Union carried out 29 more surface nuclear tests in the Ground Zero test field. This type of blast hit the Semipalatinsk region the hardest, on several occasions sending plumes laden with radioactive cesium, iodine, and strontium tens or hundreds of kilometers beyond the test site's borders. An aerial bombardment, meanwhile, contributed to global fallout: A total of 86 nuclear devices were dropped by aircraft in the direction of huge bull's-eyes etched in white chalk near Ground Zero. In these atmospheric tests, the bombs were detonated hundreds of meters above the surface. The first nuclear test moratorium in November 1958 brought a brief respite. But relations with the United States soon worsened, and within 3 years the Soviet Union had resumed atmospheric testing—with gusto. In 1962, the peak year for such detonations, 41 atmospheric blasts were carried out at Semipalatinsk and another 38 at the new range on the Arctic island of Novaya Zemlya. That proved to be the grand finale; the 1963 Nuclear Test Ban Treaty barred all testing in the atmosphere, under water, or in outer space. In all, the aboveground blasts at Semipalatinsk had released six times as much energy as the explosions at the U.S. Nevada Test Site. The action moved underground, to a warren of tunnels in the Degelen Massif, a craggy granite dome in the southern part of the Semipalatinsk Test Site. In black-and-white movies of tests in the Degelen tunnels, says Nicholas Priest, an environmental toxicologist at Middlesex University in London who has worked extensively at Semipalatinsk, “mountains would lift up, peaks would shatter.” Even today, in satellite images the mountaintops are white, but they are not snowcapped: That's rubble. The physics packages detonated in Degelen, Technical Area G, were mostly low-yield devices for testing how certain materials hold up when blitzed with radiation and devising new configurations of plutonium. Heftier detonations of up to 200 kilotons were carried out at the bottom of shafts near Lake Balapan in Technical Area B, east of Degelen. All told, 318 underground nuclear tests were conducted at Balapan and Degelen, with a further 22 tests in Technical Area M, also known as Sary-Uzen (see map). Leaving the club On 29 August 1991, exactly 42 years after the first test, Kazakhstan's president, Nursultan Nazarbayev, ordered the closure of Semipalatinsk. “We simply did not have the moral right to continue destroying our people and land with nuclear detonations,” Nazarbayev wrote in his 2001 book Epicenter of Peace (Puritan Press). But the process of decommissioning the site got off to a rough start. In the immediate aftermath of the Soviet Union's dissolution, recalls Kadyrzhanov, “the situation was quite bad at the mountain range. Local people were scavenging copper cables, collecting other scrap.” But with help from the U.S. government, Degelen's tunnels were plugged with concrete and the Balapan shafts, as well as a dozen ICBM silos there, were blown up. In the meantime, NNC officials struggled to get a handle on what precisely they had inherited. “When the Soviet Union collapsed, we had to know what happened to the test site,” Kadyrzhanov says—that is, which swaths of land were contaminated and which were clean. “Military laboratories ran a serious radioecological research program at the test site, but we received absolutely no information from them,” says Kadyrzhanov, whose institute has taken extensive radioecological measurements at the site, including aerial surveys suggesting that there is very little contamination with fission products—such as cesium-137 and strontium-90—that emit gamma rays. But although “many measurements have been made at the site, there is no comprehensive overview on all the projects conducted by a variety of different national and international organizations,” says Voigt, who's coordinating a European Commission-funded Kazakh team that by year's end expects to have assembled a database of radioecological information gathered over the years. “We want to know what the actual exposure of people living and working there is,” she says. One recent survey supports the view that vast tracts of the test site have little if any contamination. In a recently completed NATO-funded project called SEMIRAD, a team led by Priest and inorganic chemist Mukash Burkitbayev of the Al-Farabi Kazakh National University in Almaty surveyed a 550-square-kilometer wedge of land in the southeastern part of the site, stretching from the edge of the Degelen Massif to a village called Sarzhal just beyond the border. The area is particularly intriguing for two reasons. One is that on 12 August 1953 a 400-kiloton thermonuclear blast at Ground Zero sent the biggest plume of contamination ever unleashed at the site roiling southeastward, right over Sarzhal and its 2000-odd inhabitants. (According to Nazarbayev, the steppe grass glowed blue for several days after the blast.) The second feature is a pair of small crater lakes called Tel'kem 1 and Tel'kem 2: the picturesque remnants of a cluster of underground detonations aimed at testing the feasibility of using nuclear charges to excavate canals or divert rivers. As expected, the lake sediments have significant quantities of plutonium and americium. But throughout the rest of the study area, levels of two long-lasting fission products, cesium-137 and strontium-90, are generally no higher than average global fallout levels, the team notes in its February 2003 final report on the SEMIRAD project. “I was amazed how clean it was,” Priest says. He believes that most of the plume passed over the study area, depositing little radioactivity. Overall, says Kadyrzhanov, “we now believe we have a good understanding of 30% of the test site,” including the Ground Zero field, Degelen, and Balapan. Based on its incomplete picture, NNC estimates that 17% of the site has significant levels of radionuclides. “The contamination is very spotty,” says Kadyrzhanov. NNC officials hope that foreign collaborators will help pay for further surveys, which can cost several hundred dollars or more per square kilometer. “We will not be able to do it without cooperation,” says NNCDirector-General Tukhvatulin. Voigt agrees: Surveying the entire site, she says, would be “tremendously expensive.” Because it could take a decade or more to methodically map the hot spots, Kadyrzhanov says, it is premature to think about returning large swaths of land to civilian use. “It is not time yet,” he says. “There are places that may be closed for tens if not hundreds of years.” In the meantime, the devil is in the uncertainty. “Every time we conduct a survey, we find something new,” says Zhotabaev. Some of NNC's blind spots are surprising indeed. Priest, for one, has noted suspicious features on high-resolution Landsat images of the test site that, he says, “Kazakh officials would at one time deny exist.” For instance, he says, the official story was that there are only three “atomic lakes” on the test site, water-filled depressions excavated by nuclear blasts in the 1960s. “I could show you the location of at least six atomic lakes,” says Priest, adding that there are probably several more. Groundhog days The nonproliferation headache at Semipalatinsk centers on plutonium. The SEMIRAD report notes that plutonium levels in hot spots may reach 100 trillion becquerels per square meter of soil: “a potential hazard to the local population,” the report states, and “a possible source of material for terrorist ‘dirty bombs.’” The material is primarily the legacy of scores of model experiments—hydrodynamic or hydronuclear tests—conducted across the site. The aim of these experiments was to fine-tune bomb designs without having to carry out a full-fledged nuclear test. In the hydrodynamic tests, physics packages were detonated without fissile material or with it arranged in a way that the explosion would not trigger a chain reaction: for example, imploding a sheet rather than a sphere of plutonium, says Matthew Bunn, a nonproliferation expert at Harvard University's Belfer Center for Science and International Affairs. In hydronuclear tests, the plutonium released a trifling amount of nuclear energy. Shedding some light on the hydronuclear experiments, records indicate that 38 ground hydronuclear explosions—mini-mushroom clouds ranging from 250 to 1280 meters in height—were set off at Semipalatinsk in the early 1960s. But NNC officials admit that they don't have the big picture on how many model experiments were conducted and in particular how many explosions, as Priest puts it, “spread plutonium all over the place.” These were not just plutonium particles: The blasts dispersed small chunks of the metal that, after exposure to air, quickly oxidized, resulting in the patchy high levels of plutonium oxide dust in the soil. Still deeper in the shadows is a series of tests of radiological weapons conducted in the early 1950s in Area 4, north of Ground Zero. In these trials, aircraft or artillery fired munitions filled with radioactive wastes at targets. Because it was difficult to decontaminate equipment and the targets, they were simply buried on site, according to the recent memoirs of one test supervisor, Grigory Krylov. Although many other details of these obscure tests, abandoned in 1958, remain hidden, Priest and his team hope to unmask their legacy: NNC has authorized them to conduct a radioecological survey on an 800-square-kilometer section of the site covering Area 4. They are now awaiting word on funding from NATO and the U.K. government. Nearly all the Soviet bomb masters who conducted the bomb blasts, model experiments, and radiological tests at Semipalatinsk have since died or returned to a bewildering number of institutes involved in testing, taking their logbooks with them. NNC officials are hampered by a dearth of records. “The documents are not available,” says Tukhvatulin. Russia's Ministry of Atomic Energy says it doesn't have any documents. Then, Tukhvatulin says, “a document will surface somewhere.” The Russians are not prepared to relinquish some Soviet-era secrets. For example, although the Semipalatinsk site is in principle completely open to collaboration, under a Kazakh-Russian agreement on information disclosure no non-Russian entity is allowed to take measurements in a 1-kilometer-radius exclusion zone encompassing Ground Zero. Still, there are fresh hopes that more secret test sites will come to light. Through the trilateral discussions now taking place, Russian specialists have begun feeding information to their Kazakh colleagues, and Semipalatinsk has hosted a steady stream of specialists from the U.S. Defense Threat Reduction Agency, the U.S. Defense Intelligence Agency, the U.S. Department of Energy and some of its national laboratories, and the U.K.'s Ministry of Defence, all involved in sensitive nonproliferation projects at the site. Echoes of the past It must have been pleasant in Soviet times to stroll down Lenin Street, Kurchatov's main thoroughfare lined with stately trees and grand pastel buildings, with glimpses of the Irtysh River wending its way along the edge of town. “Every tree in Kurchatov had a soldier assigned to care for it,” says Oleg Pivovarov, director of NNC's Institute of Atomic Energy. In its heyday, 45,000 people, many of them soldiers, lived in Kurchatov. Today, walk a few minutes away from a clutch of administrative buildings, and in the wan twilight of a frigid spring day, Lenin Street has an eldritch feel. Many trees are dead or dying, their caretakers long gone. Most buildings are abandoned or burned out, including the shell of a stylish three-story peach-colored hotel with “1953” on its façade. The reason why much of Kurchatov today is largely a ghost town—and why a proliferation threat exists—is the withdrawal of the Russian military personnel, who once numbered four of every five residents. In Soviet times their sheer numbers and easy access to resources, such as helicopters for patrolling the perimeter, made the test site virtually impenetrable. “The Soviet Union could provide a high level of protection for the site,” says Zhantikin. “But we are a small country.” That, he says, has meant having to concentrate limited resources on surveillance at the main testing areas. But the NNC leadership now realizes how vulnerable the site is. Monitoring the movements of intruders “is impossible,” Tukhvatulin admits. Realizing that NNC must take more aggressive steps to thwart would-be terrorists, officials have asked the government for a special allotment of funds to beef up the site's defenses. Kazakh officials acknowledge that they have a long road ahead to determine just how big a threat the Semipalatinsk Test Site poses and to make it secure. Asked whether the remaining unknowns at the site worry him, Zhotabaev does not hesitate: “All the time.” 9. KAZAKHSTAN For a Long-Suffering Population, Uncertainty Reigns 1. Richard Stone SEMEY, KAZAKHSTAN—In the depths of the Cold War, a medical clinic in this city (formerly Semipalatinsk) in northeastern Kazakhstan had the ostensible purpose of caring for people with brucellosis, a bacterial disease picked up from livestock. Only after the Soviet Union unraveled did the real mission of Antibrucellosis Clinic No. 4 become clear: to observe the population near the Semipalatinsk Test Site for radiation illness and monitor radioactive contamination in the region. Interest in this exposed population continues today, and citizens benefit from the free annual checkups. Last month, a few dozen Kazakhs milling around in the clinic's waiting room had come from villages near the test site to see a doctor. Whether they have been harmed by fallout from the tests may never be known, however. Individual doses received 4 decades ago were not measured, and to calculate them with any precision so long after the fact “is almost impossible,” says neuropathologist Boris Gusev, who headed Clinic No. 4 from August 1976 until the Soviet breakup. Still, Gusev, now chief scientist for the clinic's post-Soviet incarnation—the Research Institute for Radiation Medicine and Ecology—asserts that researchers detect higher rates of some cancers and other diseases in people exposed to higher cumulative radiation doses, as estimated coarsely from contamination measurements in villages. Many experts, though, doubt that it will be possible to firm up a link. Besides the dearth of data on individual doses, there are many confounding factors—from chemical pollution to poor nutrition—that have not been adequately accounted for. “Unfortunately, we cannot establish a cause-and-effect relationship,” says Shunichi Yamashita, chair of the Atomic Bomb Disease Institute at Nagasaki University. A half-century ago, the radiation threat was real indeed. In the early days of testing, safety measures were imposed more for perception than protection. For instance, in the run-up to the first thermonuclear test in August 1953, local Communist officials were ordered to evacuate, in secret, 2250 people from Sarzhal and smaller villages. A 31 July decree ordered officials “to send comrades to villages and towns to perform public relations work to avoid possible misinterpretations.” Sarzhal villagers were allowed to return home 16 days after the blast, after radiation levels had ebbed. By the mid-1950s, however, doctors in other regions started to hear rumors of odd illnesses in the Semipalatinsk region. “The KGB prohibited us from investigating this,” recalls Saim Balmukhanov, a prominent physician and radiobiologist in Almaty. In the summer of 1957, he and a colleague were detained by military police near the test site. As they were being interrogated, a military jeep pulled up, and a man of rank came over. After listening for a while, he ordered the doctors freed. “It was the first time I met Andrei Sakharov,” the renowned physicist-cum-dissident, Balmukhanov says. Data collected by Clinic No. 4 indicated that 21 surface and atmospheric tests, the last occurring in August 1962, resulted in significant doses to people outside the test site. When Clinic No. 4 became an open institute, Gusev says, it was “a big relief” for the scientists to be able to come clean with their findings. Of the 300,000 people estimated to have received radiation doses above background, he says, some 67,000 are estimated to have received more than 1 sievert—up to 1000 times an average annual dose from natural sources and global fallout. He and others claim to have found a direct correlation between illness and radiation exposure among the highest-exposed group. Gusev's institute, up to the present day, has closely followed the health of 20,000 people in this cohort, of whom 17,000 are still living. The researchers have observed higher rates of leukemia, other cancers such as those of the lung and stomach, and heart disease, Gusev says. In a 2002 monograph, Balmukhanov and colleagues assert that the highest-exposed villagers are 40% to 50% more likely to get cancer, have higher rates of infant mortality and birth defects, and die 10 years earlier on average than nonexposed Kazakhs. The Kazakh government pushes that message hard. On the home page of the Web site of its embassy in Washington, D.C., Ambassador Kanat Saudabayev states that the nuclear blasts “caused irreparable damage to the health of more than 1.5 million Kazakhstan citizens.” The claim accompanies a plea for donations. Gusev says he's not so bothered by the exaggeration and is grateful for the money, which is used to treat people who are undisputedly ill. An open question is whether residual radioactivity threatens the local population. An intriguing connection reported between fallout and elevated mutation rates in people living near the site (Science, 8 February 2002, p. 1037)—suggesting that radiation-induced effects can be handed down from one generation to the next—has yet to be reproduced. In the meantime, a team led by Gabriele Voigt, now working for the International Atomic Energy Agency, has concluded that present exposures due to cesium-137 and strontium-90 pose no additional risk. But she says the jury is out on other radionuclides including uranium and plutonium, which may have entered the local food chain: “There are indications that these radionuclides may be found in humans,” she says. Others play down the risk. “I've told local people, ‘I don't know what doses you got in the past, but now there's no threat,’” says Nicholas Priest, an environmental toxicologist at Middlesex University in London. Yamashita's team, meanwhile, is trying to clarify the picture by conducting modern medical screening and rigorous epidemiological studies. It's a daunting mission, but Yamashita says he relishes it. Indeed, he returned to Kazakhstan earlier this week to resume the hunt for vestiges of the terrible events that are now receding quickly into the past. 10. KAZAKHSTAN Safe Haven for a Breeder's Plutonium Hoard 1. Richard Stone ALMATY, KAZAKHSTAN—In western Kazakhstan, across the Caspian Sea from Iran and Chechnya, a nuclear legacy of the Soviet era unrelated to weapons tests is causing some jitters: spent fuel from a breeder reactor that contains 3 metric tons of “ivory grade” plutonium. More than 90% of the plutonium is in the form of the isotope plutonium-239, well above the level required to make fission bombs. The cache has raised serious proliferation concerns, says nuclear physicist Dastan Eleukenov, executive director of the Center for Nonproliferation Studies office in Almaty. But he and other analysts may soon be breathing more easily. After months of negotiations, Kazakh and U.S. officials are on the verge of sealing an$80 million deal to transfer the material to a safer location: the Semipalatinsk Test Site, some 2200 kilometers to the east.

The plutonium is the product of an intriguing nuclear reactor, the BN-350 at the Mangyshlak Atomic Energy Combine (MAEK) near Aktau, a port city on the Caspian. For 25 years the liquid sodium-cooled reactor powered turbines for a water desalination plant, producing weapons-grade plutonium in the process. Kazakh officials say that the exact amount of fissile material is classified, but the U.S. Department of Energy a few years ago revealed the amount, enough for a few hundred bombs.

Getting at it would not be easy, however. The plutonium is embedded in 300 tons of highly radioactive spent fuel elements, which are encased in 2800 1-ton steel canisters and submerged in a cooling pond at MAEK. A would-be weapons pirate would have to steal several canisters and chemically extract the plutonium—“a nontrivial hassle” for a terrorist operation, says Matthew Bunn of Harvard University's Belfer Center for Science and International Affairs. But a state-run program could pull it off.

The Kazakhstan government doesn't want to take any chances, says Timur Zhantikin, chair of the country's Atomic Energy Committee. The three parties involved in the transfer—Kazakhstan, the United States, and the International Atomic Energy Agency in Vienna—held their first technical meeting last month, hashing out details on how to transport the plutonium to Semipalatinsk. In an operation expected to start next year and last until 2007, the spent fuel will be loaded into 120-ton casks and shipped by rail to Semipalatinsk, where it will be stored in underground silos at the test site's Baikal-1 reactor complex. Kazakhstan hopes to raise funds for the project at a donor conference next year, Zhantikin says.

One potential proliferation risk will remain after the fuel is gone: the MAEK brain trust that ran the reactor. “Most are nearing retirement age,” says Zhantikin, but others are expected to work on reactor decommissioning projects or find jobs connected to the booming oil fields near Aktau. “We're thinking about their future employment,” he says.

11. CLINICAL TRIALS

In Critical Condition

1. Eliot Marshall

A clinical trial to establish standards for ventilator use has been suspended for 10 months for a human subjects investigation that some researchers are calling “absurd”

It's a familiar staple of medical dramas: A patient is rushed to the emergency room fighting for breath, and when immediate measures fail to bring relief, she is placed on a ventilator. The machines have been a fixture of respiratory therapy for decades and have saved countless lives. Yet, for all this experience, consensus has been surprisingly lacking on the basics of treating patients in acute distress—such as how to set the optimal mix of volume, pressure, and oxygen. In the early 1990s, the National Institutes of Health (NIH) decided to bring science to bear on the question: It launched a $37 million, three-part clinical trial that aimed to enroll more than 2000 patients at 19 medical centers. Far from resolving uncertainties, however, the trial has sparked a bitter dispute. It has been on hold for the past 10 months after two NIH physicians claimed that its design was so flawed that it may have caused avoidable deaths. The federal Office for Human Research Protections (OHRP) has launched an investigation and is organizing its own panel to reexamine data and ethical issues in a closed session on 10 June. Much of the medical establishment seems to regard this reaction as regulatory overkill for a study that has already passed numerous ethical reviews. The American Thoracic Society, five outside consultants to NIH, and the editor of The New England Journal of Medicine (NEJM) have all disparaged the inquiry in public as “troubling”—and in private as worse—and urged that the trial be resumed. But OHRP is holding firm. “In my 21 years here, I have never seen anything like it,” says Claude Lenfant, director of the National Heart, Lung, and Blood Institute (NHLBI), which is funding the trial. Underlying this high-decibel dispute are difficult issues: how to test different treatment regimens when the standard practice itself is changing and how to design trials so that they are ethically acceptable yet provide a clear signal. The ventilator study, run by the Acute Respiratory Distress Syndrome Network (ARDSNet) and coordinated by Harvard University, consists of three linked trials. The first—a 10-center study that resulted in a highly praised report in NEJM in May 2000—sought to test whether ARDS patients fare better if a ventilator delivered large or small breaths of air. The second, completed but not published, tested different levels of oxygenation and pressure, and the third is investigating fluid management. The design of the first has been challenged, and because the second and third rely on its results, all three are at stake. According to Gordon Bernard, a pulmonologist at Vanderbilt University in Nashville, Tennessee, who chairs ARDSNet's steering committee, when the study began in 1996 “all the textbooks” recommended that patients suffering acute respiratory distress get relatively deep mechanical breaths, measured as “tidal volumes” of 10 to 15 milliliters per kilogram (ml/kg) of body weight. People believed that the lungs benefited from being well filled. But some experts disagreed, says Bernard, because “animal studies suggested that lower tidal volumes should be applied to an injured lung” and that ARDS patients would benefit if their lungs were stretched less. There was no consensus, though. Clinical practice was all over the place, says Bernard: “It was anything but standardized.” And studies indicated that up to 50% of the patients were dying. To get a crisp answer to the question of whether large or small breaths are better, ARDSNet tested two ventilator treatments on ARDS patients. By random assignment, half received a tidal volume of 6 ml/kg, a radically low number, and half got 12 ml/kg, labeled “traditional” therapy by ARDSNet. The trial stopped at 861 patients (although plans called for 1000) because a stark contrast appeared early. The 6 ml/kg group had far fewer deaths, with 31% of patients dying compared with 39.8% in the 12 ml/kg group, the study reported. “We feel we made substantial progress,” Bernard says. “All the ARDSNet investigators are using 6 ml/kg now.” But when NIH clinical care anesthesiologist Charles Natanson saw the ARDSNet data at a meeting 2 years ago, he raised an objection: The experimental therapies tested in the trial did not include what he viewed as the accepted best therapy at the time, about 8 to 9 ml/kg. Another doctor who works in NIH's clinical center, pulmonologist Peter Eichacker, agreed. (Both declined to comment for this story.) They wrote up their concerns in a detailed memo to NHLBI, which was handed up the chain of command to Lenfant. Natanson and Eichacker also co-authored with NIH statistician Steven Banks and others a critical analysis of the first ARDSNet study's design in the American Journal of Respiratory Critical Care Medicine (AJRCCM). Getting no response from NHLBI, Natanson and Eichacker filed a complaint with OHRP in July 2002. This triggered questions from OHRP, and within days, Lenfant suspended the third part of the trial for a quick review by a panel of five outside experts in August. According to ARDSNet, the design had already been looked at by at least 40 expert panels. Avoiding federal open-committee regulations, the NHLBIreviewers filed individual reports; all agreed that the design was reasonable and that the trial should resume. But OHRP denied Lenfant's request in September 2002 to restart it and instead began its own investigation in October. According to Bernard, OHRP has demanded data on more than 1600 patients. (OHRP is also looking into allegations, first raised by the Alliance for Human Research Protection in New York City, that consent procedures were inadequate because many patients were too ill to give consent themselves.) Eichacker and Natanson's concerns appear to have deepened. As they learned more about the tidal-volume experiment, according to their response to a letter published in AJRCCM, they found data indicating that most of the patients assigned to the “traditional” 12 ml/kg group would have received lower tidal volumes if they had not been enrolled in the trial. It appeared to these critics that three-quarters of the patients getting “traditional” therapy were pushed to a higher tidal volume to enhance the contrast between the two treatments under study. The clinicians' main charge was that ARDSNet's design did not incorporate the best current practice of about 8 to 9 ml/kg. They cited an authoritative 1994 textbook that warned that high tidal volumes can have “devastating effects.” The 6 ml/kg dose might have appeared beneficial only because it was contrasted with an outdated dose that was too high, they argued. The critics concluded that the evidence did not prove that 6 ml/kg was better than routine practice. Eichacker and Natanson weren't the only ones to raise questions. For example, pulmonologist Gary Salzman of the University of Missouri, Kansas City, Medical Center, voiced doubts in a letter published in NEJM before the big controversy erupted. As he says now, “We still don't know whether low tidal volume [6 ml/kg] is better than a moderate level,” because the trial didn't examine the midrange. Pulmonologist Mitchell Levy of Brown University in Providence, Rhode Island, one of NHLBI's outside experts, counters that the 12 ml/kg level was well within the normal range and well supported in the literature. If you had to adjust doses often to reflect shifting medical opinions during the course of a clinical trial, Levy says, “you could not be able to control the trial.” Levy, who calls the premise of OHRP's investigation “absurd,” warns that if OHRP buys the critics' logic, clinical researchers “could become paralyzed.” Lenfant remains baffled by the investigation. He can't understand how a clinical protocol that was so thoroughly vetted could be challenged for a bad design at such a late stage. He is particularly concerned about the “process” of OHRP's review; he doesn't know how to appeal. “It is completely unclear to all of us” why OHRP dismissed the views of the clinical experts who urged that the trial be resumed, he says. Whatever OHRP decides, clinical researchers will be able to explore the implications in detail this fall, when the American Thoracic Society plans to hold a daylong session on the controversy. 12. EUROPEAN FACILITIES Stuck in Neutron Neutral 1. Daniel Clery, 2. Gretchen Vogel* 1. With reporting by Barbara Casassus in Paris. As Japan and the United States gear up to build major new neutron sources, Europe's plans for a new facility have faltered CAMBRIDGE, U.K., AND BERLIN—A decade ago, European neutron researchers were riding high. The world's two premier sources of neutrons were in Europe, plans were moving along for a new superfacility—the European Spallation Source (ESS)—and numerous national machines catered to a thriving community. Now, however, Europe's preeminence looks threatened: Some of the older national facilities are nearing the end of their working lives, and in a huge blow last year, ESS received a poor assessment from Germany's Science Council that crippled any chance to move that facility from the drawing board and into construction any time soon. After having dominated neutron scattering for 2 decades, Europeans can only watch as the lead passes overseas. “The situation is going to become much less comfortable and more competitive compared to the U.S. and Japan,” says Bernhard Keimer of the Max Planck Institute for Solid State Research in Stuttgart, Germany. Adds Christian Vettier, science chief at the Institut Laue-Langevin (ILL) in Grenoble, France: “There is a risk that bright scientists will go to the U.S. and Japan. We need an exciting project to attract the best scientists.” But the field's proponents have not managed to excite government funders. A neutron source can be as large and costly as a front-rank telescope or particle accelerator, but it won't find the universe's missing mass or discover the Higgs boson. Neutron scattering researchers do small science on big machines. Neutron scattering “shows where atoms are and what they do,” says Robert Cywinski of the University of Leeds, U.K. Scientists shoot a beam of neutrons at a sample and register how the atoms deflect them. The x-ray photons produced by a synchrotron do essentially the same thing, but the two techniques are subtly different: Neutrons are electrically neutral, so they shoot through an atom's electron shells and ping off its nucleus. That makes neutrons much more sensitive than x-rays to hydrogen and water in biological samples, a key attribute for studying proteins and DNA. Pulsed neutrons can also provide a play-by-play as chemical reactions take place on a catalyst. And polarized neutrons—with all their spins, and hence magnetic fields, lined up—can follow magnetic excitations in a high-temperature superconductor. For these and many other reasons, scientists from a range of disciplines—from structural biology and geology to materials science and engineering—flock to neutron sources. About two-thirds of the world's 6000 neutron researchers reside in Europe. They'll often run their experiments at a small national source, then book a couple of days at a facility like the ILL or the ISIS spallation machine near Oxford, currently the top two sources in the world. Sources come in two varieties: nuclear reactors, in which neutrons are a fission byproduct, and spallation sources, in which a particle accelerator fires a beam of high-energy protons into a dense metallic target, knocking out neutrons. A turning point for neutron sources came in 1998 in the shape of a report from the Global Science Forum, a talking shop created by the Organisation for Economic Co-operation and Development (OECD) in Paris. It underlined the importance of neutron scattering for many areas of research and called upon governments to boost investment to make the most of current front-rank facilities, and for each of the major regions—North America, Europe, and Asia—to build a next-generation spallation source. In the United States and Japan, “the recommendations were taken on board and taken seriously,” says Cywinski. U.S. researchers had been starved of neutron sources for years and had suffered a major blow in 1995 when their great hope, the$2.8 billion Advanced Neutron Source, was killed off after a decade of planning (Science, 17 February 1995, p. 952). “The U.S. field was practically destroyed,” says Keimer. But by 1998, the U.S. community had pulled itself together and pushed through plans for a world beater, the Spallation Neutron Source at Oak Ridge National Laboratory in Tennessee (Science, 23 January 1998, p. 470). Japan, too, has joined the hunt; it's building a multiuse facility called J-PARC in Tokai that will include a spallation source. Both machines will begin operations within the next several years.

Europe was much slower to react to the OECD report. ILL set the ball rolling in 2001 by launching its Millennium Programme to upgrade its instruments and replace its 30-year-old neutron guides, pipes that channel neutrons from reactor to instruments. But Vettier says that ILL has had trouble persuading its funding states—France, Germany, and the United Kingdom—to fully support the $52 million project. They've given trifling amounts each year, amounting to just$10 million so far. “There is not a green light for completion,” Vettier says. The program suffered a further setback last year when French authorities assessed the reactor for seismic risk and called for major reinforcement work, improvements that will cost $24 million over 3 years. The ISIS spallation source too has been pushing for extra investment. After researchers had campaigned hard for many years for a second target station—an addition that would roughly double the number of instruments but leave the accelerator unchanged—the U.K. government announced last month that it would come up with$162 million to complete the project.

Among smaller national sources, many are approaching retirement. Denmark's DR3 reactor at Risø closed in 2000, and other smaller reactors, such as Germany's FRJ-2 at Jülich, are likely to be shut down soon. In the meantime, a funding crisis across French research could doom the Orphée reactor in Saclay. “A decision [on its future] will no doubt be taken before the end of this year,” says Pierre Monceau, director of Saclay's Léon Brillouin Laboratory. Moreover, a recent report estimated that by 2020 half of all neutron-scattering instruments now in use in Europe will no longer be operational.

View this table:

One bright spot came last month when the German government finally gave the green light to FRM-II, a controversial reactor in Garching, near Munich. Approval to fire up the reactor, completed 2 years ago, had been held up by the environment ministry, which had opposed the reactor's use of highly enriched uranium fuel. Approval was granted on the condition that FRM-II convert to medium-enriched fuel by 2010. “This event will change everything,” says FRM-II's scientific director, Winfried Petry. “Our young people will come now that they see something that has a future.”

For many European neutron researchers, however, FRM-II and upgrades at ISIS and ILL will only enable them to live on borrowed time. “The community needs a vision,” says Michael Steiner, scientific director of the Hahn-Meitner Institute in Berlin. And that, he says, is “a big spallation-type source that will open up new possibilities.” ESS was to be the embodiment of that vision. Researchers behind ESS aired their latest plans to neutron users and government officials at a meeting in Bonn in May 2002. “It was a tremendously optimistic meeting,” says Cywinski. But in a move it may long regret, the ESS council submitted its plans to the German Science Council, as two German regions were bidding to host the facility.

Last July, after assessing ESS alongside eight other large-scale projects, the Science Council said it found the proposal unconvincing. In its view, the plans failed to demonstrate that the scientific questions that ESS would tackle could not be answered using other, cheaper techniques, such as synchrotron radiation, microscopy, optical spectroscopy, or computer simulation. Researchers were “incensed,” says Cywinski. Allegations soon emerged that the council ignored parts of the analysis from its own assessment panel for ESS. Later that month, Keimer and other panel members published a letter in Nature disputing the council's final conclusions: “[The council's] statement on the cost of neutrons relative to other techniques is not based on input from us and we do not believe it would be supported by a comprehensive analysis,” they wrote.

Several months later, ESS is still stuck in neutral. A panel on neutron sources, set up by the European Strategy Forum on Research Infrastructures (ESFRI), reported last January that ESS was the way to go but not in the near term. It recommended gearing up for the facility in 10 to 20 years.

Some think it would be foolish to slacken the pace. “Planning these big facilities takes a lot of time,” says Keimer. “We need a decision now to achieve it in the medium term,” says Cywinski. But with no pan-European mechanism for funding large facilities, he says, “there is no single doorstep on which to put your proposal.”

Some optimists do see a silver lining in the limbo that Europe finds itself in. “There will be some suffering and some will go to America or Japan, but when we use the time to create a really advanced source, then we'll be doing the right thing,” says Helmut Rauch of the Technical University of Vienna. Andrew Taylor, director of ISIS, points out that the U.S. Spallation Neutron Source is essentially an ESS design dating from the 1990s. “Wouldn't it be sensible to see how they do? Why compete? Let's work with them.” Thus with their long reign at the top coming to an end, European researchers face what for some is a humbling denouement: If you can't beat 'em, join 'em.

13. ALBERT OSTERHAUS PROFILE

The Virus Collector

1. Martin Enserink

SARS. Flu. HIV. Poxviruses. Dutch virologist Albert Osterhaus is always where the action is. How many viruses can one person keep track of?

ROTTERDAM, THE NETHERLANDS—It was the study the world was waiting for. By early April, everybody suspected that a newly discovered coronavirus was the cause of severe acute respiratory syndrome (SARS). But to prove it, researchers had to infect animals with the virus, show that they got sick, and show that their bodies began producing the virus. In other words, they had to fulfill Koch's postulates.

Enter Albert Osterhaus, 54, a virologist at Erasmus University Medical Center in Rotterdam, the Netherlands. He had just what was needed: the equipment and people to infect monkeys with the virus under strict biosafety conditions and then dissect the animals to study their internal organs. The experiment started on 20 March; 3 weeks later, Osterhaus flew to Geneva to announce the results in a press conference along with Klaus Stöhr, the World Health Organization (WHO) virologist who coordinates the international SARS effort. And once again, fellow virologists were left scratching their heads and wondering: How does he do it?

Over the past 2 decades, Osterhaus has played a key role in the discovery of more than a dozen pathogenic human and animal viruses. Colleagues describe him as an ambitious, alert, and tireless workaholic with a knack for being at the right place at the right time. “He just smells viruses,” says his co-worker Guus Rimmelzwaan. Osterhaus's role in the SARS story is especially remarkable, others say, because he's also at the epicenter of efforts to control a devastating flu outbreak among poultry in the Netherlands.

Besides SARS and flu, his scientific work—a PubMed search last week turned up 498 papers—covers HIV, measles, rubella, hepatitis, herpes, hantavirus, and hemorrhagic fevers, among others. And then there are the viral diseases in rodents, African wild dogs, seals, and dolphins. “It's just completely astonishing to me,” says epidemiologist Roel Coutinho, director of the Municipal Health Service in Amsterdam. “He has only one head and two hands, right?”

His broad interests and fundraising prowess have made Osterhaus, who runs a 100-person research and diagnostic lab, an oddity in the small-scale and perennially cash-strapped Dutch research scene. In many ways, his lab resembles less an academic group than a miniature version of the U.S. Centers for Disease Control and Prevention (CDC) in Atlanta (except that CDC doesn't do Serengeti lions). Some say his strength is also his weakness—that Osterhaus is always skimming the field and never digs in deep. Still, “he has consistently done very good work,” says CDC's leading SARS scientist, Larry Anderson. “You have to admire him for his incredible energy and the things he gets done,” adds Peter Rottier of Utrecht University in the Netherlands.

The SARS monkey study was a case in point—and vintage Osterhaus, some say. Once he had received the virus from Hong Kong virologist Malik Peiris, Osterhaus decided there was no time to consult an animal ethics panel about the trial, as required by Dutch law. Instead, the director of his university's animal lab got a top official at the Dutch health department to sign off on the study, on the condition that the researchers seek approval after the fact. The move rankled animal-rights activists and a member of parliament; health secretary Clémence Ross-van Dorp later conceded it had not been handled adroitly.

But Osterhaus says he would do it again. “At that point, we were still having lengthy discussions in the WHO network about what caused SARS,” he says. “This really needed to be done as fast as possible.” He concedes he also wanted to beat rivals to the finish. He did: Nature published his paper—his 499th—about the monkey study on 15 May.

Credit fights

“Good morning! Anybody in yet? Is everybody still sleeping?” It's 9:15 on a sunny Wednesday morning, and Osterhaus, wearing a red shirt and his trademark black vest, is giving a whirlwind tour of his lab. The members of his staff look tanned but a little groggy: They just returned from a 6-day retreat in Curaçao that he held partly to reward them for their hard work. (Osterhaus joined them, but just for 2 days before he rushed back to work.) Occupying the entire 17th floor of a medical faculty building, the rooms offer breathtaking 360° views of Rotterdam's skyline. But Osterhaus doesn't care for the panorama: He hails from Rotterdam's eternal rival, Amsterdam.

Still, this port city's industrious reputation—it's said that shirts are sold in Rotterdam with the sleeves rolled up—seems to fit the lab perfectly. Ever since WHO's Stöhr invited Osterhaus to join the worldwide hunt for the SARS virus in early March, the lab has been a madhouse, says Ron Fouchier, the lead SARS researcher. Behind the united front of the WHO network, competition has been stiff, with researchers often working deep into the night. Osterhaus is a demanding lab chief, “but he works harder than anybody else,” Rimmelzwaan says.

Stöhr says he asked Osterhaus to join the WHO group because he's an “excellent scientist.” In addition to having the vital ability to do monkey experiments, another big plus was that in 2001 Osterhaus had discovered the human metapneumovirus, which causes respiratory infections in children and was an early suspect in the SARS outbreak. Osterhaus helped other labs test suspected SARS patients for the virus, which today is believed to be at best an accomplice of the coronavirus. Although mass-scale DNA sequencing is not its forte, his lab also determined the sequence of several chunks of the coronavirus genome, helping CDC complete its version of the genome and earning Osterhaus and Fouchier co-authorships on a paper published online by Science 3 weeks ago (www.sciencemag.org/cgi/content/abstract/1085952).

Just as SARS started its rampage, the vast Dutch poultry sector was struck by avian influenza. Again, Osterhaus was well positioned. In early 1997, his lab received a tissue sample from a 3-year-old boy in Hong Kong who had died from a mysterious pneumonia. Well ahead of the four global flu reference labs WHO had consulted, Osterhaus's group discovered that the virus belonged to an avian influenza strain called H5N1. At first, colleagues were skeptical, Fouchier recalls; avian influenza viruses weren't thought to cause severe disease in humans, let alone death. But other labs confirmed the finding, and later that year, 17 more people in Hong Kong were infected with H5N1 and five died. The city's entire chicken population was destroyed to quell the outbreak.

Today the Netherlands is struggling with a different strain, called H7N7, which has caused eye infections in more than 80 people and killed one veterinarian (Science, 2 May, p. 718). Flu experts worry that if the avian virus recombines with a human flu virus, it could create a lethal new strain. Along with the four reference labs, Osterhaus is working on vaccines and diagnostic tests against H7N7. “This is a dress rehearsal for a real pandemic,” Osterhaus says.

The twin outbreak has made Osterhaus a familiar face in his home country. A member of countless national and international panels, he seldom declines requests for interviews and has been on television at least 20 times since early March. “Ab loves to hear himself talk,” says Fouchier, “but he does it really well.”

Colleagues say Osterhaus's energy and personality are an asset to the bustling lab—but they can be a problem, too. Younger scientists have trouble establishing themselves under his towering ego. Osterhaus, for instance, has long insisted on being last author—a position indicating seniority—on many papers leaving his lab. His associates don't dispute his involvement in much of the work, but they need the credit. “We've fought over that, and pretty hard, too,” Fouchier says. Several compromises have been worked out; several senior researchers in his lab now grant Osterhaus last authorship on every other paper he played a role in.

Rob Gruters, who does most of the lab's work on HIV, compares his boss to Eddy Merckx, a legendary Belgian bicyclist nicknamed “The Cannibal” because his searing desire to win eclipsed a generation of his colleagues. “Ab wants it all,” says Gruters. Still, he says, Osterhaus is fun to work with. He appreciates his staff members' ideas and gives them freedom to pursue their interests—even if they're offbeat.

Deeper questions

High on a wall in Osterhaus's office hangs a poster with a picture of a sad-looking cat whose belly is the size of a soccer ball. It's suffering from feline infectious peritonitis, a fatal condition caused by a coronavirus. Osterhaus, a veterinarian by training, did his Ph.D. on the virus in the late 1970s—and never worked on coronaviruses again until SARS showed up.

But animal diseases have continued to fascinate him. As a researcher at the National Institute of Public Health and the Environment in Bilthoven, which he joined in 1978, he discovered a new herpesvirus that caused a small outbreak among seals in the Dutch Wadden sea in 1984. When the animals started dying en masse across northern Europe 4 years later, he was on the case again. At the time, many believed pollution was killing the mammals, but Osterhaus showed that a canine distemper-like virus did them in.

Wildlife viruses are still one of his passions. Many new human diseases come from animals, he says—including, most likely, SARS. His staff is currently stalking a mysterious agent that's killing red squirrels in the Netherlands.

At Erasmus University, Osterhaus has free rein to do whatever he likes. Tapping into the biodefense rage, he's started working on poxviruses. With support from the U.S. National Institute of Allergy and Infectious Diseases, Gruters is working to develop an HIV vaccine. Osterhaus has two companies to commercialize his findings; most of the revenues are cycled back into the lab, he says, where they help pay for the wildlife studies or, for that matter, last month's Caribbean junket. Last year, he licensed the worldwide rights to develop vaccines and antibodies to the human metapneumovirus to MedImmune, a biotech in Gaithersburg, Maryland. The down payment alone reeled in $10 million—most of which goes to the university. And yet he's not satisfied. Osterhaus missed out on several recently discovered fatal viruses—such as the Nipah virus from Malaysia—because he doesn't have a biosafety level 4 lab, the multimillion-dollar pressure-controlled facilities where scientists work in space suits. He wants the medical center to help him build one, and with all the money he's bringing in, he reckons they owe him. “I simply need it,” he says. “Otherwise I can't move ahead.” Some colleagues, especially in the small world of Dutch virology, charge that Osterhaus is a scientific stamp collector. But others say such complaints are motivated by professional envy. “I think he's the best and most fun researcher we have in Holland,” says virologist Jaap Goudsmit, chief scientific officer of Crucell, a Dutch biotech. In any case, the criticism doesn't seem to bother Osterhaus. He can't imagine spending his career on just a few viruses—it would be boring, he says. Besides, there are deeper questions that drive him: Where do new diseases come from? What causes them to pop up at a certain place and time? And how can they be stopped? “Maybe I should sit down more to write about that,” he says. “Often there's just not enough time.” And then he has to dash off. He needs to prepare for a major grant review meeting the next day and has to be in a TV studio at 6 p.m. to talk about SARS. And, of course, there are new viruses to hunt down. The 500th paper is just weeks way. 14. MEDICAL RESEARCH COUNCIL Out of the Frying Pan Into the MRC 1. Jocelyn Kaiser Colin Blakemore emerged unscathed from a decade of threats from animal-rights extremists; will his fortitude help him restore the MRC's reputation? The past 3 years have brought peace, of sorts, for Colin Blakemore. He no longer has a police escort when he gives a seminar, for instance, nor do demonstrators regularly picket outside his home. But although the threats and mobs have subsided, Blakemore still varies his route to work. “I will never get out of the habit of looking under my car for bombs,” he says. For his defense of animal research, including his own experiments on kittens, activists for more than a decade made his life a living hell. The calluses that Blakemore, 58, has developed while fighting those battles may offer some protection in his new job: head of the Medical Research Council (MRC), the U.K. government's main biomedical research agency. In recent months, MRC has been fiercely criticized for allegedly making poor funding decisions and losing touch with the scientific community. The agency has taken on “a corporate mentality,” charges developmental geneticist Robin Lovell-Badge of MRC's National Institute for Medical Research (NIMR) in London. Lovell-Badge and others believe that Blakemore, who now heads MRC's Centre for Cognitive Neuroscience at Oxford University, is the right person to turn things around. He's “an excellent choice” for an institution “in deep trouble,” says neurobiologist Andrew Lumsden of King's College London. Even those who think MRC is in good shape are happy with Blakemore's appointment: “He's a strong character, dynamic, and will take the MRC to the next phase,” says Aaron Klug of MRC's Laboratory of Molecular Biology in Cambridge. As a young scientist Blakemore hit the ground running, earning a Ph.D. at the age of just 23 and moving on to an area of neuroscience pioneered by Nobelists David Hubel and Torsten Wiesel: how experience shapes the developing visual system. In experiments with cats and monkeys, Blakemore has explored the extent to which connections among neurons in the visual cortex are fixed by early stimulation—work that has helped doctors understand how children develop amblyopia, or “lazy eye.” He also proved precocious at explaining science to the public, at 32 becoming the youngest ever commentator for the Reith lectures on BBC Radio. He has since appeared on hundreds of TV and radio programs. But Blakemore's public embrace took a sinister turn in 1987, when an increasingly radical animal-rights movement targeted him because of his work on kittens. He was beaten, his house vandalized, his daughters threatened with kidnapping. Two bombs were mailed to his home, including a mailing tube packed with needles that was brought inside by his children. (They didn't open it.) Blakemore did not cower: He routinely met with the demonstrators and co-founded a group to promote discussion between scientists and activists (Science, 4 June 1999, p. 1604). “He argues the case so articulately, the more moderate groups go away,” says Lumsden. And in the past few years, many of the extremists have been locked up. Blakemore's private life may have grown calmer, but that hasn't stopped him from courting controversy. Even some of his admirers say they were taken aback when in April 2002, after Israel sent troops into the West Bank, Blakemore joined 124 other scientists on a letter calling for European agencies to rule Israeli scientists ineligible for grants. Blakemore insists that the letter was “by no means a call for a boycott” of Israeli scientists, and last January he co-authored a commentary in Nature proposing criteria for when a boycott would be warranted. One prominent scientist says that Blakemore has “tied himself in semantic knots” defending his position. A spirit of engagement, nonetheless, ought to serve Blakemore well at the beleaguered MRC. The$700 million agency supports 41 institutes, centers, and units, and it awards fellowships and grants to scientists at universities. In March, a parliamentary committee accused MRC of mismanaging finances and failing to adequately consult the scientific community on plans for a large new population study, the U.K. Biobank (Science, 28 March, p. 1958).

One widespread concern that Blakemore recognizes is that under CEO George Radda, whose 7-year term ends in September, MRC changed the criteria for small grants to require that applicants be part of large collaborative groups. The rationale is to compel scientists in various fields to work together and share expensive equipment. But some scientists—especially young ones and those in fields not requiring big collaborations—have struggled to find partners for proposals. “Single investigators have been driven away from the MRC,” Blakemore says, to other sources of funding. Besides reviewing whether to restore the single-investigator grants, Blakemore hopes to “even out” recent fluctuations in funds that have meant that some top-rated grant proposals were turned down.

Blakemore is less critical than some of his colleagues, however, of plans for Biobank, an attempt to link the genetic makeup of 500,000 middle-aged Britons with lifestyle factors, such as smoking and exercise, and disease. The government's portion of funding for the planned \$70 million project, Blakemore points out, is new, earmarked money that is not cutting into funds for research grants. Biobank would “take advantage of the almost unique quality of [British health] data,” he asserts.

Another “hot potato,” Blakemore says, is a proposal by MRC's leadership to move its largest institute—NIMR in Mill Hill in north London—to Cambridge and halve its 760-member staff by 2010. Many at Mill Hill “are extremely upset” by the plan, Blakemore says. However, he sympathizes with the rationale behind the move: embedding institutes in universities to foster interaction between academic researchers and clinicians. The MRC council will respond to feedback on the proposal in July, well before Blakemore takes up the reins.

Although he will have plenty to keep him busy at MRC headquarters in London, Blakemore expects to maintain his lab and spend at least a half-day a week there. That sets him apart from past MRC administrators, who tended to move on from doing research after assuming the post. But for Blakemore, staying rooted in a scientific community that he has endured tribulations to defend ought to be, by now, second nature.