# News this Week

Science  02 Aug 2002:
Vol. 297, Issue 5582, pp. 748
1. DRUG-ABUSE RESEARCH

# White House Stirs Interest in Brain-Imaging Initiative

1. Andrew Lawler

BOSTON—Thanks to drug-war money, Massachusetts General Hospital just dedicated one of the most sophisticated magnetic resonance imaging machines around. With a whole-body magnet and a 460-ton steel shield, the multimillion-dollar device promises to give researchers new insight into the brain circuitry of drug craving and euphoria. The device is one of dozens of neuroimaging machines planned for labs across the country, all paid for by White House drug czar John Walters's office. Their debut is creating a euphoria of its own among researchers. They are putting together a major initiative to use the machines in concert with advances in genetics and animal research to vault the field into the forefront of neuroscience—what Walters calls “a bold new model for drug-abuse research.”

The idea for a brain-imaging initiative—seen as a decade-long, $100-million-plus effort—is generating excitement among policy-makers such as Walters, officials at the National Institute on Drug Abuse (NIDA), and academic researchers, many of whom gathered here last month at a White House- sponsored meeting on reducing drug demand. By gathering data from thousands of human subjects, they hope to understand the genetic and physiological underpinnings of drug abuse. But skeptics warn that the ambitious effort must be carefully designed, lest it produce reams of unusable data—and damage neuroimaging's already mixed reputation. Researchers involved in the effort say that the technology is ready: “I'll do everything I can to kick this off,” says Albert Brandenstein, chief scientist for the Counterdrug Technology Assessment Center (CTAC), which is part of the White House Office of National Drug Control Policy. “We want the best people in the world—and our carrot is the best equipment in the world.” The U.S. government is spending nearly$19 billion this year on treating and preventing drug abuse and interdicting illicit drugs; the vast majority goes for law enforcement. NIDA, part of the National Institutes of Health, has a $933 million budget, but that money pays for research, not infrastructure. So Brandenstein's office stepped in a few years ago to provide imaging facilities and technical support;$14 million is allocated for 2002. Almost a dozen machines are completed or under construction—including three at NIDA—and CTAC hopes to fund 40 to 50 over the next 5 years, officials say. This growing network would form the basis for the proposed initiative.

Mass General neuroscientists Hans Breiter and Greg Gasic will submit a grant proposal to Brandenstein's office this month for a pilot program, the first step toward a national program. CTAC would pay for the pilot project; NIDA would then take over to run a full research initiative. “This would be no smaller than the genome project, once full blown,” says Gasic, who foresees “a concerted effort by many institutions.” The Mass General imaging center, which houses the new machine, is a cooperative venture with nearby Harvard and the Massachusetts Institute of Technology (MIT).

The Mass General team envisions a three-phase effort. A 1-year pilot project would recruit more than 100 subjects—mostly siblings—divided into four groups: nicotine addicts, cocaine addicts, depressives, and a healthy control group. Using the imager, researchers would test their reactions to a handful of exercises involving perception of beauty, monetary rewards, and pain. To date, the largest neuroimaging studies have been limited to about 50 people. Phase two of the Mass General effort, lasting 3 to 4 years, would recruit up to 4000 people from the New England region and collect images and DNA from each. The third phase would fund multiple centers in a national effort involving an estimated 8000 to 20,000 human subjects. The costs could be about $2.5 million for the pilot program—borne by Brandenstein's office—and could top$100 million in NIDA funding for phase three, according to officials familiar with the long-term plans.

The first two phases—if approved by peer reviewers at CTAC—would test whether links could be established between genetics and neuroimaging data. “Two years of work will give us the confidence to step forward to see if there is enough data for a major initiative,” says Brandenstein. Ultimately, “a goal of this effort is to develop the genotypes and phenotypes of the addicted human brain,” which will lay the foundation for better treatment and prevention programs, adds Richard Millstein, NIDA's deputy director. NIDA likely would farm out work on the third phase to a variety of labs. Such an interdisciplinary collaborative effort, insists Walters, “is the best hope for a better future in the struggle against drugs.”

Many researchers are enthusiastic. “This sounds wonderful; we've learned a great deal with longitudinal studies,” says Eric Kandel, a Nobel laureate and neuroscientist at Columbia University in New York City. “This could be a Framingham study on mental health,” referring to a long-term examination of heart disease in that Massachusetts town. Others, however, are cautious about promising too much. “You may end up with data that is uninterpretable,” warns Nora Volkow, a neuroscientist at Brookhaven National Laboratory in Upton, New York. “There's a consensus [that] we need to do our homework before embarking on such a program.” She says that brain images from the same person can vary from day to day, and others note that drug users typically use several substances, complicating efforts to sort out variables.

The genetics portion will pose its own challenges, researchers say, particularly attempts to link specific genes with physiological changes. “Sure, we should try,” says Harvard provost and neuroscientist Steven Hyman. “But we need to have the greatest humility as to where we are.” Many agree, however, that neuroimaging and genetics demand collaboration. “There is no question about the importance of these two forces in understanding cognitive issues,” says Phil Sharp, who heads MIT's McGovern Institute for Brain Research. The challenge, he adds, is to come up with studies that can pass muster in peer review, be replicated, and build up large databases for future researchers.

Breiter acknowledges that neuroimaging has had a reputation for producing “pretty pictures” but not replicable data. “It has been characterized as pseudocolor phrenology, but thanks to very rigorous animal neuroscience, we know how [neural] circuits work.” Responding to colleagues, he revamped his proposal to include a slow scale-up. “The worst-case scenario,” he says, is that it would end with the pilot studies, giving “a neural circuitry-based picture of nicotine and cocaine use and depression.”

Such a picture would aid drug-abuse research, which Harvard University psychiatrist Perry Renshaw says has long suffered from a lack of good clinical studies using neuroimaging. Larger studies, he says, hold the key to making use of the technology's possibilities. They also require more money and are certain to raise controversial issues about confidentiality, gender, and ethnicity. “If this is not well thought out, it will hurt us,” adds Volkow. But given new facilities, funding, and the strong support of the White House, drug-abuse researchers might have a good shot at riding to the forefront of neuroscience.

2. BIOTERRORISM

# A Call for Restraint on Biological Data

1. Jennifer Couzin

Two events last week are prompting a public debate over a hot-button issue that has quietly been discussed in the scientific community since last fall's anthrax attacks: Should unclassified research that might conceivably help bioterrorists be openly published?

A handful of members of Congress filed a resolution criticizing the publication by Science of a paper on poliovirus and calling on journals, scientists, and funding agencies to take more care about releasing such information. Separately, the American Society for Microbiology (ASM), which represents 40,000 scientists, sent a letter to the National Academy of Sciences (NAS) 22 July requesting a meeting of biomedical publishers to discuss whether and how to publish research that might be co-opted by terrorists. NAS plans a meeting this fall.

The Science paper, published online 11 July (www.sciencemag.org/cgi/content/abstract/1072266), describes the assembly of poliovirus from stretches of DNA obtained by mail from specialty reagent suppliers. The publication troubled Representative Dave Weldon (R-FL). Along with seven other Republicans, Weldon introduced a resolution 26 July criticizing Science's publisher, the American Association for the Advancement of Science (AAAS), for publishing “a blueprint that could conceivably enable terrorists to inexpensively create human pathogens.” The resolution, which has been referred to three congressional committees, also calls on government funding agencies to reconsider how they classify research. The polio study, funded by the Department of Defense and led by Eckard Wimmer at the State University of New York, Stony Brook, was unclassified.

Many microbiologists say that they see no threat to national security in the polio paper, because the virus's DNA sequence is available over the Internet and techniques for building it have long been known. “The colleagues that I have spoken with … do not feel there was any information presented in the publication that was national security information,” says Andrew Onderdonk, a microbiologist at Harvard and editor-in-chief of the Journal of Clinical Microbiology, one of 11 journals published by ASM. At the same time, some biologists have condemned the publication for needlessly raising public fears (see Letters, p. 769).

Alan Leshner, AAAS's chief executive, defended the decision to publish: “The technique reported in Science is neither a practical nor efficient method for making more complex, lethal viruses,” he said, noting that methods used in this research had been previously published and that the virus Wimmer's group produced is less virulent than natural poliovirus.

Weldon's call for rethinking open publication of potentially sensitive microbial work, however, echoes months of conversation among biologists. “Everyone is walking on eggs,” says Sam Kaplan, chair of the ASM publications board. Editors of the ASM journals, which frequently publish research on dangerous pathogens, have been mulling over policies since last December. The Department of Defense, meanwhile, is funding a workshop of journal editors and national security experts on 12 August in Washington, D.C., on the publication of research it funds that is potentially related to biological warfare.

Absent clear guidance, some scientists are taking matters into their own hands. Ronald Atlas, president of ASM, says that the group's journals have received “a dozen or two dozen inquiries” from scientists afraid to publish their work in full. ASM's answer: Incomplete papers are not eligible for publication. In at least one case, though, a gutted paper did slip through: a report on smallpox sent to the Journal of Clinical Microbiology by Thomas Smith, director of the Virology Laboratory at the Mayo Clinic in Rochester, Minnesota, and colleagues at the Centers for Disease Control and Prevention in Atlanta, Georgia.

The Mayo-funded report described a way to rapidly identify smallpox by a small segment of its genetic sequence. After the paper was accepted, Smith says, federal employees he declines to identify raised concerns. They worried that a terrorist could alter this bit of sequence to slow identification of the virus during an attack. Smith's group agreed to remove critical details, and the journal published the shorter version in June. Experiences like this, according to Atlas, drove the society to call for the NAS publishers' meeting.

Even though scientists agree that some research results might be risky to release—and that they might not know what constitutes a security threat—they are wary of suppressing data. Furthermore, some say, biodefense research is needed now more than ever, and keeping it secret will only make fighting terrorism tougher. “You can dream up all sorts of extreme scenarios on how bioterrorists can benefit from information,” says Paul Keim, an anthrax researcher at Northern Arizona University in Flagstaff. But suppressing information “will hurt our effort to combat bioterrorism.”

Scientists might have to live with some censorship, however, says Claire Fraser, director of The Institute for Genomic Research in Rockville, Maryland: “There could be more harm than good done by publishing a paper,” she thinks. “That's going to be very hard for scientists to deal with.”

3. BIOTERRORISM

# Student Charged With Possessing Anthrax

1. David Malakoff

A University of Connecticut graduate student has become the first researcher charged under new antiterrorism laws with mishandling a potential bioterror agent. Federal prosecutors last week charged Tomas Foral, 26, with unlawfully possessing anthrax-tainted cow tissue.

Foral can avoid a trial—and up to 10 years in prison if convicted—by completing a community service program. But the young scientist is upset by the charge, which he says he can't afford to fight and believes is the result of “a misunderstanding” with a laboratory superior.

The case highlights the increasingly treacherous legal landscape surrounding pathogen research, some researchers say. “I fear this young man has gotten caught up in an overreaction to [last year's] anthrax attacks,” says Ronald Atlas, a bioterrorism expert at the University of Louisville, Kentucky, and president of the American Society for Microbiology.

Foral's troubles began late last October at the university's pathobiology laboratory in Storrs, Connecticut, where he is a master's degree student working to develop a detection test for West Nile virus. After a professor asked him to help clean out a malfunctioning basement freezer, Foral found a rusty container labeled “anthrax” holding about a half-dozen vials of cow tissue collected in the 1960s. Foral says that after a brief conversation with the instructor, he saved two of the vials in another locked laboratory freezer for future research. According to Foral, the instructor was unclear about what to do with the vials, so Foral froze them. (Science could not reach the instructor for comment.)

One month later, following an anonymous tip, police investigating an anthrax death in a town about 100 kilometers away came searching for the vials. After Foral turned them over on 27 November, the lab building was closed for more than a week. FBI agents began an investigation, including searches of Foral's home and university room, where they photographed textbooks and journal reprints, he says.

On 22 July, U.S. Attorney John Danaher announced that the government was charging Foral with possessing a controlled biological agent in violation of the USA Patriot Act, an antiterrorism law rushed through Congress last October (Science, 2 November 2001, p. 971). Foral was not covered by any of the law's exemptions, such as possessing anthrax for “bona fide research” purposes, prosecutors said in a statement.

Foral can avoid prosecution by doing community service, continuing to cooperate with investigators, and staying on the right side of the law. Prosecutors emphasize that his participation would not be “evidence of guilt.” But Foral says he is deeply disheartened by the ordeal and worried that it might harm his efforts to get into medical school. The Czech-born American citizen, who serves in the National Guard, says the case has already caused his name to be added to an immigration watch list: When his military unit reentered the United States after training in the Caribbean, he notes, he was delayed for hours while FBI officials checked out his story. “It's gotten Kafkaesque,” he says.

University officials, meanwhile, have watched with concern as Foral's case has unfolded. Some schools, such as the Massachusetts Institute of Technology in Cambridge, had already hinted that the criminal sanctions and security requirements imposed by the Patriot Act and the more recent bioterrorism law (Science, 31 May, p. 1585) might force them to end research on regulated agents such as anthrax. “Many researchers are still unaware of these laws,” says Atlas. “Deans are terrified,” he adds, that one of their students could be next.

4. PSYCHOLOGY

# Violent Effects of Abuse Tied to Gene

Some children who suffer physical, sexual, or emotional abuse become violent adults. But many do not. Now a new study of both genetics and social surroundings points to the influence of a particular genotype on aggressive behavior in young adults from a troubled background.

On page 851, a team led by clinical psychologists Terrie Moffitt and Avshalom Caspi, both of King's College London and the University of Wisconsin, Madison, reports that a certain form of a gene that breaks down neurotransmitters makes men more likely to be violent, but only if they were maltreated as children. “This is a very important piece of work,” says geneticist Greg Carey of the University of Colorado, Boulder. “It's pretty convincing for just a single study.”

The gene codes for an enzyme called monoamine oxidase A (MAOA), which metabolizes several kinds of neurotransmitters in the brain. By getting rid of excess neurotransmitters, MAOA helps keep communication between neurons functioning smoothly. Studies of lab animals show that knocking out the MAOA gene makes adult mice more aggressive. The first suggested evidence in humans came from a 1993 report of a Dutch family (Science, 18 June 1993, p. 1722). Several men in this family had a defective MAOA gene—none of the enzyme was found in their cerebrospinal fluid—and were prone to impulsive bouts of aggression. But because the mutation is extremely rare, no one has replicated the finding in other families.

To see whether the MAOA gene influences aggressive behavior in the broader population, Moffitt and Caspi's team turned to New Zealand's Dunedin Multidisciplinary Health and Development Study. The study, begun in 1972, has followed 1037 children since birth. Hoping to get as homogeneous a genetic background as possible, Moffitt and Caspi selected 442 subjects with four white grandparents. “It's about as refined as it can be,” Moffitt says.

As expected, the team discovered that severely maltreated boys were more likely to exhibit so-called antisocial behavior than boys who had suffered little or no abuse. But the researchers also found that antisocial behavior was more likely in males with the genotype for low MAOA activity who had been mistreated. The 55 boys in this group were about twice as likely to have been diagnosed with conduct disorder in adolesence as the 99 mistreated boys with the high-activity genotype. And they were three times more likely to be convicted of a violent crime by age 26. Although the 55 males who had experienced moderate or severe maltreatment and also had the low-activity genotype made up only 12% of the study group, they committed 44% of the crimes. “They're doing four times their share of rape, robbery, and assault,” Moffitt says.

But environmental influences were critical, Moffitt found. In the absence of abuse, having the low-activity genotype didn't make boys any more likely to be antisocial. Jon Beckwith of Harvard Medical School in Boston agrees, although he'd like to see the finding replicated: “I would use this as a wonderful class example of how social factors can play an enormous role in expression of behavioral traits.” Moffitt views the results as an example of how accounting for environmental factors can help reveal a gene: “Finding the stressor can be a magic key.”

There are caveats. The link between the MAOA alleles and the activity of the enzyme in these males is only inferred, Beckwith points out. Also potentially confounding the study is that antisocial behaviors might depend on social situations, not just genes, adds sociologist Troy Duster of New York University.

Replicating the results will be important, researchers say, although this might be easier than in previous studies because the sample was drawn from the general population. Confirmation could also lead to better intervention strategies. Social workers and therapists would benefit from knowing which abused kids are most at risk, notes criminologist Alfred Blumstein of Carnegie Mellon University in Pittsburgh.

Legal implications are less clear. Although some attorneys might argue that the MAOA genetic defect results in diminished capacity, Hal Edgar of Columbia Law School in New York City doesn't think judges will buy it. “This particular study in and of itself is not going to shape the [legal] culture,” he says. And experts warn that it's much too early to discuss whether drugs might counter the effects of low MAOA activity.

Experts also say that it's important to remember that many genes probably influence violence and other antisocial behaviors. Or as Carey says, the strongest genetic marker for violence is still the presence of a Y (male) chromosome.

5. NEUROSCIENCE

# Long-Awaited Technique Spots Alzheimer's Toxin

1. Laura Helmuth

STOCKHOLM—Alzheimer's disease is notoriously difficult to diagnose, particularly as it begins to take hold. Researchers suspect that therapies, when they become available, will work best if given early, however, raising the need for a test that spots the first signs of this dementia-causing disease. On 24 July at the International Conference of Alzheimer's Disease and Related Disorders here, a team revealed the first images from a positron emission tomography (PET) technique that picks up one of the defining—and first—features of Alzheimer's disease.

“People are going to point to this particular presentation and say, ‘This is when we started making progress’” on visualizing Alzheimer's disease, says Mark Mintun of Washington University Medical Center in St. Louis, Missouri. This putative marker, as well as others reported at the meeting, could be invaluable not only for diagnosis but also in clinical research, conference attendees say.

Clinicians settle on a diagnosis of Alzheimer's disease through a process of elimination. Pencil-and-paper tests of memory and problem solving, reports from family members, and standard brain scans suggest that Alzheimer's disease is the culprit, but researchers can't be entirely sure until an autopsy reveals brain tissue riddled with senile plaques and neurofibrillary tangles. The main ingredient in plaques is a protein called β amyloid, which congregates in the brain well before symptoms of dementia appear and is thought to initiate most of the damage.

Several teams have been racing to find a nonintrusive way to visualize β amyloid in a living brain. William Klunk, Chester Mathis, and colleagues at the University of Pittsburgh spent about 10 years searching for a molecule that fits the bill. Their best candidate, dubbed PIB, crosses the blood-brain barrier, isn't toxic, and can be clipped to a radioactive tag to light up plaques, animal studies have shown.

Starting in February, a group led by Henry Engler of Uppsala University in Sweden injected the molecule into nine people with symptoms of “mild” Alzheimer's disease and five healthy controls and then used a PET scan to see where it went. At the conference, images of the results audibly took the audience's breath away: In healthy people, the marker sailed right through the brain and was well on its way to exiting via the central fluid-filled ventricles. But in people with early Alzheimer's disease, the marker stuck in the cortex, particularly in the frontal lobes and temporal-parietal areas, two of the brain regions most damaged in the disease.

“This is something we've all been waiting for,” says Michael Pontecorvo of Mitsubishi Pharma America in Warren, New Jersey. Adds Randy Buckner of Washington University, “this opens a new window on what we think to be the primary marker associated with disease.”

β amyloid can be detected in a far more unexpected place, another team reported. Lee Goldstein of Brigham and Women's Hospital in Boston and colleagues have found that the protein also collects in the lens of the eye. What's more, in a postmortem study of eyes from 16 elderly donors, half of whom suffered from Alzheimer's disease, all of the patients had a rare type of cataract on the edges of the lens-apparently caused by β-amyloid deposits. But this doesn't mean that an eye exam for Alzheimer's disease is around the corner, cautions Goldstein, in part because it's not yet clear how early in the disease's progression the protein might become visible in the lens.

Standard brain scans, if analyzed carefully, can also show signs of impending Alzheimer's disease, reported Nick Fox of the National Hospital for Neurology and Neurosurgery in London. By comparing magnetic resonance images taken 1 year apart, Fox and his colleagues found that healthy elderly people lose about 0.2% of their brain volume each year. People who start out with minor memory complaints and progress to Alzheimer's disease, on the other hand, lose about 2.8% a year. This dramatic shrinking is evident in the hippocampus, which helps store memories, and the frontal and temporal lobes.

The technique doesn't provide a quick diagnosis: It takes time to get a clear picture of Alzheimer-like decline. But it could help monitor patients' progress in clinical trials, says Fox. Their performance on tests of memory and problem solving varies a lot from day to day, making it tough to tease out any potential improvements. By providing a more accurate endpoint, he suggests, brain scans might enable researchers to make clinical trials “smaller and faster.”

6. ASTROPHYSICS

# 'Winged' Galaxies Point to Black Hole Mergers

1. Robert Irion

X literally may mark the spot as astrophysicists hunt for colliding black holes. Results of a new mathematical model, published online this week by Science (http://www.sciencemag.org/cgi/content/abstract/1074688, maintain that cross-shaped radio galaxies harbor massive black holes that suddenly flipped their spins, probably by absorbing black holes from other galaxies. When combined with a census of these distinctive galaxies, the model suggests that such titanic encounters happen about once a year in the cosmos.

Observations of galactic cores, including the center of our Milky Way, have revealed that most galaxies host supermassive black holes with millions or billions of times the mass of the sun. In active galaxies—those that spew enormous fountains of energy into space—theory holds that these vortexes spin at awesome rates as they devour gas and stars. The incoming matter spirals into a raging disk, which shoots jets into space at nearly the speed of light. Astrophysicists don't yet understand this process, but they assume that the jets mark a black hole's spin axis.

Previous surveys showed that about 7% of active radio galaxies have X-shaped, or “winged,” jets ranging in shape from narrow beams to cones. Astronomers thought these features pointed to a precession of the central black hole, much as Earth's spin axis wobbles over time. However, recent high-resolution radio images of some winged galaxies show sharp breaks where a pair of jets angles off into a new direction, rather than sweeping out gradual curves (see figure, above). “That's clearly not precession,” says astrophysicist David Merritt of Rutgers University, Piscataway, New Jersey. “It has to flip over.”

The likeliest mechanism is the arrival of a second massive black hole during a galaxy collision, say Merritt and his colleague, radio astronomer Ron Ekers of the Australia Telescope National Facility in Sydney. According to their model, an incoming black hole with at least 20% of the mass of its partner will knock the main black hole off kilter, no matter how rapidly it spins.

The calculation agrees with an independent analysis of black hole mergers using Einstein's theory of general relativity, says astrophysicist Scott Hughes of the University of California, Santa Barbara. “It's really hard to torque a black hole around by a large amount without having something as massive as another black hole slam into it,” Hughes says. He and astrophysicist Roger Blandford of the California Institute of Technology in Pasadena are preparing their work for publication.

From estimates of how long the X-shaped radio lobes persist, Merritt and Ekers deduce that a typical large galaxy will undergo a black hole-tilting crash once every billion years. That's enough for one such event to pop off somewhere in the universe each year. The result bodes well for astrophysicists who hope to observe the intense ripples in spacetime, called gravitational waves, that should cascade from such mergers.

The research should spur theorists to figure out how gigantic black holes manage to merge—instead of forming binaries that waltz for billions of years, as most models hold. “This suggests that nature does find a way to bring some black holes together,” Merritt says. “We're just not sure how.”

7. EUROPEAN PATENTS

# Tough Stance on Stem Cell, DNA Claims

1. Gretchen Vogel

BERLIN—Biotech players hoping to stake claims on human stem cells or DNA sequences in Europe saw a couple of warning shots whiz across their bows last week. On 24 July, the European Patent Office (EPO) strongly limited a controversial patent covering stem cell technology, striking out all references to human or animal embryonic stem (ES) cells. And the influential Nuffield Council on Bioethics, a British think tank, called on patent offices around the world to refrain from awarding patents on DNA sequences.

EPO cautioned against reading too much into a single decision in the rapidly developing field of stem cell research. “One could not possibly deduce a patent policy from a single case,” says EPO spokesperson Rainer Osterwalder. Nevertheless, EPO's stance contrasts sharply with policies at the U.S. Patent and Trademark Office (USPTO), which has granted half a dozen patents involving human ES cells, including a broad patent on the technique used to derive cell lines. That patent's owner, the Wisconsin Alumni Research Foundation, claims that its patent covers all import and use of human ES cells in the United States. Its application for a similar patent in Europe is under review at EPO.

Last week's EPO ruling concerned the so-called Edinburgh patent, which covers techniques for using molecular markers to identify stem cells. Granted in 1999 to developmental geneticist Austin Smith of the University of Edinburgh and Peter Mountford of Stem Cell Sciences in Melbourne, Australia, the patent generated controversy 2 years ago when Greenpeace charged that its transgenic animals claims could be construed as covering the creation of transgenic humans (Science, 3 March 2000, p. 1567).

EPO admitted that it had erred in allowing that claim, explaining that the examiner had simply overlooked the possibility that the patent might cover the creation of human beings. And the patent holders said they never intended to conduct such experiments. Indeed, when the storm broke they proposed modifying their claim accordingly. But before they could do so, 14 parties, including Greenpeace and the governments of Italy, Germany, and the Netherlands, filed opposition petitions.

After 3 days of hearings, an EPO review panel concluded that the patent conflicted with the European Patent Convention, which governs EPO, on two grounds: The convention prohibits patents involving the use of human embryos for industrial or commercial purposes, and it requires that work described in a patent be specific enough to be repeated by an expert in the field. The claims involving ES cells were too vague, the panel said, in part because the patent application was filed in 1994, several years before scientists first reported isolating human ES cells. Faced with that ruling, the patent holders agreed to strike all references to ES cells, leaving only claims dealing with stem cells derived from adults or fetal tissue. The panel allowed the narrower patent to stand and will issue a written decision within several months, after which either side can appeal.

Some experts argue that the ruling does not preclude future patents on the use of human ES cells. According to George Schlich, a patent attorney for the University of Edinburgh and Stem Cell Sciences, researchers might still be able to win European patents on processes involving differentiation of human ES cells into tissues that could be used to treat diseases such as diabetes or heart disease. The key, he says, would be to focus a patent on the resulting tissue rather than the starting material, whether that is ES cells or stem cells derived from adults.

In the meantime, if an international panel gets its way, patents on DNA sequences could become harder to win. A day before EPO's ruling, the Nuffield Council on Bioethics issued a report recommending that patents on DNA sequences be “the exception rather than the norm.” The report calls for patent offices to distinguish among different uses of a genetic sequence—for example, a specific genetic test, a method of gene therapy, or production of a therapeutic protein—and in general to grant patents on a specified use rather than on the DNA sequence itself. Making such distinctions could help clear up some of the ethical and legal debates over DNA patents, says biotechnology patent expert Mildred Cho of Stanford University. But implementing the Nuffield Council's recommendations, she says, “would require a major shift in thinking at the USPTO” and other patent offices.

8. 2003 U.S. BUDGET

# NSF Gets Big Lift; Pluto Mission Backed

1. Jeffrey Mervis*
1. With additional reporting by Jocelyn Kaiser and Andrew Lawler.

Senators Barbara Mikulski (D-MD) and Kit Bond (R-MO) have delivered on their promise to put the National Science Foundation (NSF) budget on a 5-year doubling track. But they also served notice that they are putting the agency on a tight leash.

Mikulski and Bond are chair and ranking member, respectively, of a Senate Appropriations Committee panel that has written a $91 billion bill covering the 2003 budgets of NSF, NASA, the Environmental Protection Agency (EPA) and dozens of other agencies. Last week, the full committee approved a 12% increase for NSF, to$5.35 billion, the largest percentage boost for any major agency in its jurisdiction. The legislators also overrode a White House plan to halt work on a Pluto mission and gave EPA science programs a slight increase.

The committee's vote is only the first step in a budget journey that might not conclude until after the November elections, but it's a big push for NSF's supporters, who have been urging Congress to match the explosive growth enjoyed by the National Institutes of Health in the past 4 years. “This is definitely a positive signal,” says Samuel Rankin of the American Mathematical Society and the Coalition for National Science Funding, which is aiming for a 15% increase.

NSF's increase, for the fiscal year starting 1 October, is more than twice the 5% boost requested by the president. Research programs would jump by 15%, to $4.13 billion. Big winners within that account would include the physical sciences, graduate student stipends, a program to help have-not research states, cybersecurity, and research instrumentation. At the same time, Mikulski and Bond would throttle back on a new program to improve math and science education (Science, 11 January, p. 265), expressing concern about NSF's ability to spend its$160 million allotment for this year. And they want to keep a closer eye on NSF's management of big projects. In addition to delaying the start of a proposed $12 million network of environmental monitoring stations, the bill would cut$15 million from a new $35 million earthquake detection and research network, called EarthScope, and freeze the money until NSF hires a permanent director to oversee big new research facilities (Science, 12 July, p. 183). The legislators also gave a whopping 28% boost to the agency's in-house watchdog, the inspector general, to carry out more audits of NSF programs. EPA's science and technology account would receive a 1.7% boost (not counting$90 million in supplemental funding in 2002) to $710 million, more than reversing a 4% cut that the White House requested. The increase includes$10 million for the STAR graduate fellowship program, which the president had proposed transferring to NSF without providing funding—effectively killing it. The Senate bill restores the money and keeps the program at EPA. In a related move, the NSF portion of the bill deletes the proposed transfer of $76 million in programs from EPA and two other agencies (Science, 8 February, p. 954). NASA's$15 billion request—just slightly more than the current budget—would increase by $200 million in the Senate bill. Legislators also set aside$105 million for a Pluto mission that the White House has put on hold, an amount that falls $50 million short of what mission planners say is needed to keep it on track for a 2006 launch. Legislators also restored a$7 million cut proposed by the White House in the $17 million National Space Biomedical Research Institute in Houston, a move that had angered Texas lawmakers. But the NASA portion of the bill is chockablock with projects, such as$2 million for an aquarium in Maine, that benefit the districts of specific lawmakers but are not related to NASA's mission. The list of so-called earmarks is expected to grow this fall when the House marks up its version of the bill.

9. WOMEN'S HEALTH

# U.K. Hormone Trial to Pause for Review

1. Martin Enserink

For at least 3 months, no new patients will be enrolled in a large trial of hormone replacement therapy (HRT) taking place in the U.K., Australia, and New Zealand. The U.K. Medical Research Council (MRC), the trial's main sponsor, ordered the pause last week and decided to ask an international panel to recommend whether to continue the trial in the face of evidence that prompted termination of a similar U.S. study 3 weeks ago. However, women already enrolled will be asked to keep taking their pills.

Safety reviewers halted the U.S. study, designed to test the long-term benefits and risks of HRT, after an interim analysis found that taking a combination of estrogen and progestin was too risky. The reviewers concluded that an increased risk of breast cancer, stroke, and heart disease outweighed benefits related to colorectal cancer and bone fractures (Science, 19 July, p. 325).

Despite the findings of excess risk, U.K. leaders of the Women's International Study of long Duration Oestrogen after Menopause (WISDOM) saw no compelling reason to halt their own trial. Both WISDOM's steering committee and an independent safety panel unanimously concluded that the U.S. study, part of the Women's Health Initiative (WHI), had not conclusively demonstrated the increased risk of heart disease. That meant the balance of risk and harm from HRT was still uncertain, they said, and it was ethical to keep enrolling women, provided they were fully informed about the risks (Science, 26 July, p. 492).

But the MRC Council, meeting 26 July, decided that an international group of experts should study the matter and issue a recommendation by October. “The MRC has a duty to safeguard the well-being of the women who have volunteered for the WISDOM study,” MRC chief George Radda said in a statement. “We must be absolutely satisfied that it is right to press ahead … with the study.”

“It sounds like a very reasonable approach to me,” says Stanford University's Marcia Stefanick, principal investigator of WHI. Unlike the WISDOM leaders, Stefanick says she has no doubts about the WHI results, although she notes that researchers differ in what they regard as convincing. But WISDOM steering committee chair Rory Collins worries that recruitment in the trial, already behind schedule, will suffer, even if it resumes after October: “Once you stop a tanker, it's quite difficult to get it going again.”

10. HIGH-ENERGY PHYSICS

# Tevatron Sees Light at End of Tunnel?

1. Charles Seife

Six months ago, the Tevatron accelerator was in trouble, plagued by technical woes that threatened to cripple the device's research program. Now scientists working on the project say things are looking up—mostly. Scientific results from the accelerator are beginning to trickle out, and a 2-week shutdown in June might have marked a turning point in the battle against the machine's problems. But the accelerator's particle beams still fall far short of their target brightness levels, and scientists still await the Tevatron's reemergence as a flagship accelerator.

A year and a half ago, the Tevatron, which smashes protons and antiprotons together at enormous energies, began operating again after a $260 million refit. Despite months of tinkering, however, scientists and engineers couldn't boost the beam's luminosity—its “brightness”—high enough to begin the bulk of the accelerator's research program. By January, Tevatron scientists had devised a plan for attacking the accelerator's problems and had set performance benchmarks for the rest of the year (Science, 8 February, p. 942). Six months later, the accelerator has missed every single luminosity benchmark. “We're still a factor of 2 short,” says Stephen Holmes, head of the beams division at Fermi National Accelerator Laboratory (Fermilab) in Batavia, Illinois. A major problem with the accelerator lies in the system that accumulates, accelerates, and stores antiprotons—which, unlike protons, are hard to produce. Fully 80% of the antiprotons were supposed to survive the trip from the accumulator system to the collider, but in January, a mere 30% made the journey intact. “Really, until April, we had no idea what the physical cause” of this problem was, says Holmes. So, despite Fermilab's best efforts, “we topped out at about 40%. We were pretty much stuck.” In April, however, scientists at Fermilab figured out that the antiproton problem was caused by intrabeam scattering. “When the antiprotons are going around and around in the antiproton accumulator, they are confined to a very small space, and they are bouncing off each other,” says Holmes. “This tends to heat the beam, making it get bigger. It wants to blow up.” Scientists had anticipated problems, but this effect was worse than expected. Now a 2-week shutdown in June might have solved the antiproton problem, Holmes says. While the accelerator was turned off, engineers improved the beam cooling system and refocused the magnetic optics that keeps the beam tight. Now about 50% to 60% of the antiprotons survive the trip to the accelerator, and the number is rising. With that roadblock removed, last week the Tevatron's luminosity surged to a record-setting 2.64 × 1031 inverse square centimeters per second—still far short of the benchmarks but a vast improvement. “The problem seems like it's moving downstream, out of the antiproton facility,” says Holmes. Physicists now have enough data to start doing real science with the Tevatron, says John Womersley, spokesperson for the D0 experiment on the Tevatron. “There are a lot of interesting things we can do with a small data set,” he says. Teams have already used Tevatron data to measure properties of the W and Z bosons, carriers of the weak nuclear force, at an energy at which these properties had not been measured before. The amount of data already collected should roughly double the number of top-quark sightings, Womersley says. “We're beginning to do physics,” he says, “but this doesn't mean we're happy to run like this for years.” Tevatron engineers still have several other problems to solve. One particularly sticky one arises from the unwanted mutual influence of the proton and antiproton beams, which travel close to each other in the accelerator. This beam-beam interaction is more severe than expected at low energies, and engineers are going to tear out a bottleneck in the accelerator where the beams get too close together. That makes physicists at CERN, the European laboratory for particle physics near Geneva, nervous, as their next big machine, the Large Hadron Collider (LHC), also has two high- luminosity beams in close proximity. “For the LHC, we are pretty confident that [beam-beam interactions] are not going to be a problem, but anything can happen,” says Steve Myers, Holmes's counterpart at CERN. Partly for that reason, CERN is flying a handful of its scientists to Fermilab to assist their American colleagues. “The main effort is to help them out, and in so doing, we get hands-on experience dealing with problems that might affect us,” says Myers. “In repayment, when we start running the LHC, they will help us. This is the deal of the collaboration.” According to Mike Harrison, head of the superconducting magnets division at Brookhaven National Laboratory in Upton, New York, the Tevatron's fundamentals are sound: The performance at top energy is good, and the antiproton production system works—it's just a matter of getting all the parts to work together, usually a struggle when getting an accelerator up and running. If Tevatron scientists manage to bring their accelerator's performance up to its design expectations, they will be ready for the next big challenge: adding a system to recycle used antiprotons. With luck, the Tevatron's greatest difficulties lie behind it rather than ahead of it. 11. MOLECULAR BIOLOGY # In Yeast, Prions' Killer Image Doesn't Apply 1. Jennifer Couzin Researchers studying misfolded proteins have found prions in yeast and fungus; unlike those in mammals, they don't seem to harm the host Twenty years ago, a curious new agent emerged from obscurity to join the cast of biology's villains. Dubbed prions, these infectious proteins have been fingered for causing an array of rare but horrific brain illnesses. They are suspected of triggering “mad cow disease,” for example, which is thought to have crossed the species barrier and killed more than 100 people in Europe. Prions' fearsome reputation is enhanced by mystery: Researchers have not been able to determine exactly how they do their dirty work. And although they are the prime suspect in several diseases, they haven't been experimentally proven to be the sole cause of any. But researchers have identified one feature that all prions appear to share: a pernicious shape. Whereas many proteins bend into alternate forms and still function properly, the hallmark of human prion proteins is that they have morphed from a normal, harmless protein into a contortion that seems to turn them deadly. And having made that change, according to prion researchers, they become unrelenting bullies, forcing other proteins to adopt their misfolded shape. This standard picture is shifting, however. A radical new line of inquiry is casting human prions not as prototypical evildoers but more like the rare bad guys in an extended family of eccentrics. Among the prions' mild-mannered siblings and cousins, according to studies published since 1994, are certain prionlike agents in yeast and fungus. Like their malevolent kin, they appear to cause other proteins to adopt their shape, yet their hosts usually seem to suffer little or no harm. Some research even indicates that these prions might perform useful functions such as helping cells survive in tough environments or transmitting beneficial qualities from one generation of cells to the next. Although controversial, these ideas are gaining support. “We have to get out of the mindset of considering a prion as a disease [agent],” says Susan Liebman, a molecular geneticist at the University of Illinois, Chicago. Even if the new prions do turn out to be benign, they might still be useful models for researchers studying human diseases. Increasingly, biomedical scientists are turning to colleagues in the yeast and fungus fields for help in understanding how proteins fold, misfold, and possibly trigger brain lesions. For example, Alzheimer's disease and prions are both associated with distinctively arranged clusters of misfolded protein, called amyloids. In addition, a number of other neurodegenerative diseases, including Parkinson's and Huntington's diseases, are marked by comparable clumps of misfolded protein in the brain. The amyloids of Alzheimer's and the protein clusters of Parkinson's are not considered transmissible from one cell to another, but similar mechanisms might be behind all the protein misfolding in these structures. ## Surprising finds As early as the 1950s, before prions were identified, researchers had noted bizarre behavior displayed by some strains of yeast and fungus. French scientists observed that heritable, non-DNA elements of certain fungi could be passed on to their offspring. In the 1970s, another French group found a similar anomaly in yeast: An inherited element that was not produced by a gene seemed to appear spontaneously. “That at first seemed very strange to me,” says Reed Wickner, a geneticist at the National Institute of Diabetes and Digestive and Kidney Diseases in Bethesda, Maryland. In 1989, he began considering parallels between the odd yeast traits and those of a human prion, then a relatively new term referring to the misfolded version of the mammalian protein PrP, the normal form of which is especially prevalent in nerve cells. Wickner focused on two heritable, non-DNA elements: [URE3] and another that had been described in the literature, [PSI+] (Prion names are normally bracketed.) Both had already been found to be transmissible from one yeast cell to another in cytoplasm—similar to viruses. Wickner proposed in Science in 1994 that the pair represented “self-perpetuating” versions of two normal yeast proteins, Ure2p and Sup35p, respectively (Science, 22 April 1994, p. 566). Somehow, these proteins could spontaneously convert to the [URE3] and [PSI+] forms, which in turn converted other normal proteins around them—just as the prion form of PrP is thought to do in human brains as it destroys them. This would explain why genes didn't produce these proteins: As prions, they appeared only when the normal protein form misfolded. Neither [URE3] nor [PSI+] had an obviously remarkable effect on yeast, though. [URE3] alters the cell's nitrogen metabolism, and [PSI+] allows the synthesis of certain proteins to continue when it would normally be stopped. Despite the apparent absence of disease, Wickner labeled the pair prions; at the time, they were the only proteins other than [PrP] to garner that title. Yeast researchers went on an extended dig to confirm Wickner's theory and uncover additional prions. Biologists examined [URE3] and [PSI+] for similarities, which they hoped would guide them. One critical feature quickly emerged: Both yeast prions contained unusually long stretches of two amino acids, glutamine and asparagine. Some believe that these boost the chance that proteins will form certain flat structures, called β sheets, which likely enable protein-protein interactions. Another link turned up in studies by a team including Liebman, Yury Chernoff, now a yeast prion expert at the Georgia Institute of Technology in Atlanta, and Susan Lindquist, a yeast geneticist then at the University of Chicago and now director of the Whitehead Institute at the Massachusetts Institute of Technology. They found that at least one yeast prion, [PSI+], seemed to require the presence of a so-called heat shock protein, HSP104, to maintain its prion structure. Heat shock proteins are molecules that protect a cell from stress and guide protein folding. Since these researchers reported on HSP104 and [PSI+]'s relationship 7 years ago, Lindquist and others have found that heat shock proteins play a role in both formation and inhibition of additional yeast prions (see profile). ## Metaphysical questions The hunt is on for new prions, and labs are stalking them in different ways. One popular approach involves removing the rich glutamine-asparagine portion of a known prion protein—considered the “active” bit that permits conversion—and replacing it with a chunk of candidate protein that is suspiciously prionlike. Researchers can then test whether the revamped protein still contorts into a prion. Another test is to flood a cell with a suspected prion protein, and then determine whether the normal form of the protein loses its standard function. With these and other tactics, researchers have pinpointed at least four prions in yeast and are debating the properties of several more. They have also found a candidate prion in a species of fungus. Lindquist predicts that an additional 20 yeast prions await discovery, a conviction supported by a paper Liebman published in Cell last summer. While screening to find the gene behind a possible prion, Liebman hit on 11 that passed the test. Two were known or suspected of generating prion proteins; the rest all produced proteins rich in glutamine and asparagine and, she concluded, might potentially be responsible for still-undiscovered prions in yeast. “We know that these sequences [rich in glutamine and asparagine] are preserved and are very prone to forming prions,” says Jonathan Weissman, a cell biologist and biochemist at the University of California, San Francisco. “Exactly why they're there is much harder to answer.” Such teleological quandaries are the subject of polite disagreement. Lindquist reported in Nature in 2000 that almost half the time, yeast cells containing the [PSI+] prion react differently to their environment. In more than a quarter of these cases the effect is positive, such as tolerance of harsh conditions. She hypothesizes that in certain situations the ability to “switch on” prions is beneficial and that these self-perpetuating proteins play a major role in guiding yeast evolution. But others aren't so sure. Skeptics point out, for example, that although yeast prions are abundant in laboratory strains, they're almost unheard of in the wild. (The strain commonly used in the lab and the source of much prion data, Saccharomyces cerevisiae, is itself difficult to unearth in a natural environment.) Mick Tuite, a molecular biologist at the University of Kent at Canterbury, U.K., has tested 15 yeast strains taken from AIDS patients and found the [PSI+] prion in none of them; a couple of strains, though, have yielded another yeast prion, [RNQ+]. A fungal prion, [Het-s], isolated from Podospora anserina, has been found to exist naturally, says Sven Saupe, a geneticist at the University of Bordeaux in France. This oddly folded protein appears to inhibit Podospora organisms from fusing with one another—a widespread phenomenon among certain fungi. This might help prevent the spread of infection, Saupe says, but it's not clear why a prion would be selected for this role. ## What makes a murderer? As researchers such as Lindquist and Saupe struggle to outline the role prions might perform in yeast and fungi, they keep running into perhaps the most vexing question of all: Why don't prions such as [URE3] and [Het-s] kill the way [PrP] does? That mystery is bringing yeast and fungi experts together with those who study [PrP]—still the only mammalian prion known—and certain diseases marked by amyloids and misfolded proteins. View this table: Huntington's disease is the target of one such collaboration. Michael Sherman, a biochemist at Boston University School of Medicine, saw yeast as a tool for investigating what makes the culprit protein, called huntingtin, toxic. Huntingtin damages yeast cells, and varying levels of toxicity can easily be measured in this system. To his surprise, the yeast prion [RNQ+] seemed to increase huntingtin's ability to do damage. Sherman teamed up with Chernoff, the Georgia Tech yeast prion expert. The pair determined that converting [RNQ+] to its normal protein shape prevents huntingtin from aggregating and killing the yeast cells. This suggested that huntingtin alone isn't sufficient to launch disease in yeast. And in the 10 June issue of The Journal of Cell Biology, the group hints that still-undiscovered prions or prionlike proteins in humans might also be critical to forcing huntingtin to aggregate. “The question is whether there's something similar in mammalian cells,” says Sherman. “There are probably many, many other prion-type proteins … but we don't know of them.” Lindquist champions the view that prionlike proteins are common in mammals, including humans, but that they might not normally cause disease. “It depends entirely on the kinds of proteins they interact with,” she says. It's also possible that some prions are intrinsically more prone to toxicity than others. Despite their potential for harm, Lindquist adds, prions likely extend benefits, too. “It's a wonderful means for very stably transmitting information,” she says, referring to the prion's ability to convert proteins in cells around it to the same form. “Once you set up certain states, having structures that tend to be self-perpetuating makes a lot of sense.” The logic is finding support in at least one provocative new line of inquiry. Tantalizing evidence for benign prionlike proteins—they don't match up to true prions—is coming from Nobelist Eric Kandel's lab at Columbia University in New York City. Kandel and lab member Kausik Si are studying a common protein in neurons called CPEB, a section of which resembles parts of prions in yeast. Preliminary evidence suggests that the protein can self-perpetuate in mammalian brains, the pair reported at a National Academy of Sciences meeting in March. Although cautioning that the evidence is extremely preliminary, they speculate that it might play a role in storing information—in other words, in memory. Lindquist argues that further study of prions—or self-perpetuating proteins, as she likes to call them—will help explain what makes at least one of them harmful. Hazardous and not, she and others believe, many more prions are out there, waiting to be revealed. 12. MOLECULAR BIOLOGY # Susan Lindquist: Prion Expert Leads the Whitehead Institute 1. Eliot Marshall Susan Lindquist likes to say she “stumbled into” a career in science. She doesn't mean she hesitated. It's just that she never planned to be a member of the National Academy of Sciences, a leader of a fast-developing area of protein studies (see accompanying story), or the director of a major academic center: the Whitehead Institute for Biomedical Research in Cambridge, Massachusetts, which she was chosen to head last fall. As Lindquist tells it, her strategy was pretty simple: to follow her curiosity. “As a kid,” Lindquist said in an interview earlier this year, “one of my favorite activities was to make ‘mixtures,’” concoctions of berries and “weird things” gathered in the neighborhood and fermented under plastic wrap “to see what would happen.” Looking back, she sees this “enticing” aspect of discovery as drawing her into science, the “inventiveness … of poking at something and having an idea.” Her Swedish- and Italian-American parents did “not have a chance to go to college,” Lindquist says, and she was not programmed to be an academic superstar. During her adolescence in Chicago in the 1960s, “becoming a housewife was pretty much the program. … That's what a woman was supposed to do.” But she was attracted to medicine as an undergraduate at the University of Illinois and got hooked on biology one summer working in the lab of biologist Jan Drake. She was accepted into Harvard's graduate school in 1971 and wound up in the high-powered lab of biochemist Matthew Meselson, her thesis adviser, later famous as a MacArthur award “genius” and expert on biological warfare. Even then, she didn't entertain the “expectation that I would ever have a lab of my own.” After all, she says, there were hardly any women on the faculty. Today, Lindquist runs not just a lab but an entire institute, and she will need more than curiosity as she takes on some difficult management challenges. She took over from Gerald Fink, a yeast genetics expert like herself, who remains on the faculty. The change came at a critical time. The Whitehead, which is marking its 20th anniversary this year, has been a pioneer in basic research but is perhaps best known now for its contribution to sequencing the human genome. Led by molecular biologist Eric Lander, the mass-production effort boosted the Whitehead's profile—and its income. In fact, the genome center's income—mostly federal—accounts for 71% of the Whitehead's$140 million budget. But as sequencing winds down, Lander reportedly has been exploring ways to spin off his evolving interests in a new center (Science, 21 December 2001, p. 2451). Such a spinoff, which would leave a hole in the institute's finances and faculty, has yet to materialize. But if it never does, this could be difficult, too, for the genome group and the rest of Whitehead might need to sort out their overlapping, postsequencing agendas.

Steering the Whitehead through these changes is Lindquist's job. Neither she nor Lander wants to discuss it. The topic of the genome center's future, a Whitehead spokesperson explained, is “off the table” for public discussion. Nor is Lindquist prepared to comment on other strategic issues that came up at an institute retreat earlier this year—such as how to bring more youthful leaders into the faculty, which is top-heavy with Whitehead founders.

Lindquist is forthcoming, however, when she talks about her work, which includes prions: misfolded proteins suspected of triggering a variety of brain disorders including “mad cow disease.” She got into this field, she says, through her graduate research on heat shock proteins (HSPs), a category of agents also called “chaperones” because they guide other proteins within cells and help fold them into the proper shapes.

Lindquist chose HSPs as a thesis topic on her own, because Meselson was “distracted” with weapons-control activities at the time, she recalls. After spending “2 weeks” in the library, she decided to investigate proteins produced by fruit fly larvae in heat stress, the biochemical products associated with heat shock “puffs” visible on fly chromosomes then being studied by a Harvard faculty member, Sarah Elgin.

Now a professor at Washington University in St. Louis, Missouri, Elgin says her own work on chromatin benefited when Lindquist and another member of the Meselson lab, Steven Henikoff, characterized the heat shock proteins and made it possible to identify some of the genes involved. Elgin thinks Lindquist has an “intuition for what's interesting” and is “fearless about going in new directions.” Meselson doesn't recall Lindquist's struggle to find a thesis topic, but he does recall her work: “Susan was very careful to make herself familiar with all aspects of what she was doing; she focused on important problems—and she has not changed.”

Lindquist moved to the University of Chicago in 1976 and stayed for more than 2 decades, working with yeast to explore heat shock proteins, protein folding, and, later, prions and diseases that might involve protein folding. Among the lab's achievements, she says, was showing that the alternative shapes a prion takes correlate with its biological properties. It was “quite hard” to leave Chicago, she says, because she had lived there “all my life, and most of my family is there.” But the lure of the Whitehead—and the rich scientific infrastructure of Cambridge—were too much to resist.

Although Lindquist is reluctant to talk about management decisions, she says she launched a salary review that significantly raised the Whitehead's pay for postdocs. This had “complex” repercussions, she notes, but she adds that she would like to go further in “reducing hierarchy.” She has set up a lab of her own at the Whitehead and will investigate links between protein folding and human disease. For the institute, her main goal is to support individual excellence: “I want to … allow people a broad scope and palette and let them go off in different directions.” More specific plans will become clear, according to a spokesperson, after an outside consulting group delivers a “top-to-bottom strategic review,” due in September.

13. SENSING

# Brainstorming Their Way to an Imaging Revolution

1. Daniel Clery

In June, a handpicked team of researchers locked themselves away in an R&D hothouse to produce a new detector of elusive terahertz waves. Their prototype is already being tested

OXFORD, U.K.—Terahertz waves penetrate fog, peer through paper and clothes, and look into human tissue, but their useful properties are terra incognita to most because of the huge cost of existing sensors. Last week, however, a team of scientists from across Europe began testing an imaging chip that could open up this long-neglected part of the electromagnetic spectrum to new applications, from medical imaging to satellite observations of Earth. The device itself is intriguing enough, but equally novel is how it's being developed.

The process began in November 1999, when a pair of physical scientists embarked on a breakneck effort to fabricate a new material that can completely block out terahertz waves. This radiation, in the nether region between infrared and radio waves, is hard to detect, but a so-called photonic bandgap material impervious to terahertz waves could revolutionize imaging devices, greatly improving their ability to peer through materials opaque to light of many other wavelengths. Chris Mann of Rutherford Appleton Laboratory (RAL) near Oxford, U.K., and Ramón Gonzalo of the Public University of Navarra in Spain cloistered themselves away in RAL for a month to come up with the goods.

The duo succeeded, producing a prototype terahertz-blocking silicon material. Musing over their accomplishment in a Pamplona bar in May 2000, Mann, Gonzalo, and Peter de Maagt of the European Space Agency (ESA) agreed that this kind of forced, intense teamwork—a mini-Manhattan Project approach—might be just the ticket to take the next, more difficult, step in terahertz imaging. They hatched a plan that night to assemble a crack R&D team to design a terahertz imager that could be deployed in space and elsewhere.

Two years later, the Star Tiger project, funded by ESA, is yielding its first fruits. A team of 11 researchers from across Europe, under the leadership of de Maagt, Mann, and RAL colleague Geoff McBride, last week began putting a prototype terahertz imager through its paces at an RAL lab. The scientists still face big technical hurdles if they are to reach their goal: production of a much more sophisticated chip by the end of the project in October. But they are zealous converts to the agency's novel multidisciplinary team approach. “If you want to find something innovative, it's the best way,” says Luisa Deias, an electronics engineer from Italy and the team's sole female member. “This isn't work,” adds British materials scientist James O'Neill. “We're just having fun.”

The seed that germinated in the Spanish bar 2 years ago fell on fertile ground at ESA. Back at the agency's technology research center in Noordwijk, the Netherlands, de Maagt mentioned the idea to his superiors. It quickly moved up the hierarchy, and in April 2001, ESA launched a feasibility study into building a terahertz imaging chip. Six months later, Mann and de Maagt got the go-ahead for a $650,000 project. All objects emit terahertz waves, just as they emit infrared radiation, but terahertz waves are much harder to detect. Existing imaging devices were originally designed by the military to help land aircraft in fog, but they are complex, bulky, and expensive. Chip-sized detectors could be mass-produced and thus could open up new markets. A terahertz imager at an airport, for example, would be able to see through passengers' clothes and reveal hidden weapons, which emit more terahertz waves than the human body does. Every airliner could have a detector in its nose cone, allowing the pilot, on a foggy night, to see the runway. And—the reason ESA got involved in the project—satellites could use them to look down at Earth through cloud cover or up at the stars at this little-studied wavelength. To make a tiny chip-sized detector practical, you have to ensure that as many of the weak incoming waves as possible make it into the detector rather than leaking into the surroundings or into the chip material itself. The semiconductors that chip substrates are normally made of are a big impediment: They soak up terahertz waves like a sponge. That's where photonic bandgap materials come in (see sidebar). This substrate reflects rather than absorbs the waves, and it can be shaped like a cone to funnel incoming waves to a chip's detector elements. It's like making a telescope mirror out of polished glass rather than black felt. “Photonic bandgap materials have the potential to revolutionize terahertz technology,” says Mann. First, however, the Star Tiger organizers had to recruit a team. They advertised the project earlier this year, and in late April, 16 candidates—ranging from newly minted graduates to seasoned postdocs—gathered in a country house in Oxfordshire, U.K., for an intense weekend of personality tests, technical presentations, group discussions, and interviews. “Everyone was really stressed,” says Dutch team member Frank van de Water, a waveguide designer. The lucky 11 were informed of their selection a few days later. But when they arrived at RAL on 5 June, they found an empty room with computers and other equipment still in boxes. “Setting up the room was a team-building exercise,” says McBride. The team building continued when Mann whisked the whole group off to Cornwall—his home territory—for 2 days of brainstorming, with occasional breaks for surfing and other fun. Several weeks into the project, in late June, spirits remained high. “The magic thing about Star Tiger is putting everyone in one room,” says British space scientist Alec McCalden. The result, team members say, is an intellectual ferment rarely found in an academic setting. “Here everyone has very different points of view. It opens your mind,” says Spain's Iñigo Ederra, an antenna engineer. Adds his compatriot and fellow antenna designer Jorge Teniente Vallinas, “Every day there is a new idea from someone.” Star Tiger's leaders ensure that the scientists are not distracted by administration. “Without the burden of bureaucracy, we will be able to do in 4 months what would normally take 1 to 2 years,” says Deias. Preparing the prototype this month has been the group's first real test. As a demonstration of their ideas, they have constructed a single-pixel detector made up of a cone-shaped feedhorn channeling radiation to oscillators, mixers, amplifiers, and detectors embedded on a chip below. The aim is to produce a terahertz image of a human hand in 30 seconds or less. To do this with a single pixel, moving mirrors focus radiation from different parts of the hand onto the pixel in quick succession. Ultimately the team wants to build as many as 32 pixels onto a chip, either in a square array or in a row that can be scanned across the target. They hope to do away with the moving mirrors by electronically steering the array so that it is sensitive to waves coming from different directions. The team has a good shot at succeeding, says Don Arnone, chief executive of TeraView, a Cambridge, U.K., firm set up last year to make terahertz medical imaging systems. Such devices illuminate the body with a source of terahertz waves and analyze reflected signals for signs of cancer, as the waves are better than x-rays at contrasting cancerous and healthy areas in the skin, breast, or other soft tissues. A system like Star Tiger's—which picks up only naturally emitted waves—would be “commercially very attractive” and have many applications, particularly in airport security, Arnone says. The key challenge is creating an imager that can scan a whole person fast enough. The October finish line for Star Tiger might seem a distant target for the researchers now working full out. “Enthusiasm will flag. People get tired,” says McBride. “But they look after each other. They will look after the ones that fall.” And team members are convinced that the experience will benefit them down the road. According to German materials scientist Alfred Zinn, “it will pay off in the future when we go off to our universities and one day can call one another up and say, ‘I've got this crazy idea, will it work?’ The trust is already there.” 14. SENSING # Terahertz on a Chip 1. Daniel Clery The Star Tiger team refers affectionately to its photonic bandgap material as a “woodpile.” In the bottom layer, silicon “logs” are lined up in one direction, while the next layer has them perpendicular to the first, and so on. This arrangement creates a macroscopic version of a crystal lattice: a structure with arrays of holes like the serried ranks of atoms in a crystal. And just as a semiconductor's crystalline structure can forbid the movement of electrons with particular energies—an energy bandgap—a silicon woodpile blocks certain wavelengths of radiation. The size and spacing of the holes are finely tuned so that when radiation of a particular wavelength range impinges on the material, the waves are refracted and reflected by all the surfaces around the holes to the point where they shift out of phase and cancel each other out. The waves can't propagate through it, so the material behaves like a three-dimensional mirror: Shine light from any direction, and it gets bounced back out. To make Star Tiger's terahertz “mirror,” a dicing saw carves grooves out of both sides of a thin silicon wafer, forming a two-layer woodpile. Stacking produces thicker piles. Because the technology doesn't exist to detect terahertz waves directly, the detector circuits of Star Tiger's chip generate terahertz waves of a known frequency and mix them with incoming terahertz waves funneled down a feedhorn cut in a top layer of woodpile. The two sets of waves cancel each other out in some places and reinforce in others to produce a wave with a much lower frequency, equal to the difference between the two original frequencies. This manageable radio-frequency signal is filtered, amplified, and detected by other chip elements, then sent to a computer for analysis. Building a working detector is within reach, says Dutch waveguide designer Frank van de Water. “This goal is definitely achievable,” he says. “We have to believe in it.” 15. CHEMICAL WEAPONS # U.S. Research on Sedatives in Combat Sets Off Alarms 1. Alexander Stone* 1. Alexander Stone writes from New York City. U.S.-funded studies on how to turn such drugs as Valium or Prozac into weapons undermine a treaty against chemical weapons, critics say Drugs such as Valium and Prozac might seem like the antithesis of modern weapons, but not to the U.S. government, which is sponsoring research into the feasibility of combat use of sedatives and other drugs that inhibit the function of the central nervous system (CNS). The work, described in documents obtained by Science, is part of a broader effort to create an arsenal of nonlethal weapons for soldiers and police. But critics say that turning such drugs into tools to subdue hostile forces would run counter to an international treaty that bans the use of chemical weapons. Although the United States ratified the Chemical Weapons Convention in 1997, it maintains the right to use nonlethal riot-control agents in law enforcement and certain combat situations, despite objections from other countries. Indeed, the government's interest in nonlethal weapons has grown significantly over the past decade as U.S. forces have been deployed in such urban settings as Somalia, Kosovo, and Bosnia, says a spokesperson for the U.S. Marine Corps, which oversees the Department of Defense's (DOD's) Joint Non-Lethal Weapons Program (JNLWP). Funding for studies of nonlethal weapons has jumped from$14 million in 1997 to $36 million in 2001. A domestic program aimed at giving law enforcement officials better ways to resolve a variety of situations, from dispersing rioters to rescuing hostages, is also under way. Work on the use of drugs as nonlethal agents is being conducted at the Institute for Emerging Defense Technologies of Pennsylvania State University, University Park. Created in 1997 to research nonlethal weapons for the Marine Corps, the institute is supported in part by a contract worth up to$42 million from the Corps to the university, and its director, engineer and retired Col. Andrew Mazzara, was formerly head of JNLWP.

Mazzara and engineer John Kenny are currently carrying out a study that tries to gauge the effects on humans of breathing in an aerosolized mixture of calmatives (substances that depress or inhibit CNS function and produce tranquil or calm behavior) and pepper spray—a commonly used crowd-control agent. The study, funded by the National Institute for Justice (NIJ), uses high-tech dummies to monitor absorption rates, concentration, and flow of the mixture into the bloodstream and various organs. Mazzara says that NIJ, the research branch of the Department of Justice (DOJ) and a member of JNLWP, asked him to do the study “because they see violent reactions to OC [pepper] spray.” The study, due to be completed this year, hopes to identify the optimal dosages needed to temporarily subdue targets.

The research builds on a 2000 review paper by Kenny and two colleagues at the institute that urged the Marine Corps to give “immediate consideration” to weaponizing sedatives such as diazepam (Valium) and selective serotonin reuptake inhibitors such as fluoxetine (Prozac) and sertraline (Zoloft). The scientists also proposed research into the possible weaponization of “drugs of abuse” and convulsants such as those commonly found in rat poison. Kenny, who leads JNLWP's human effects advisory panel, says the Corps did not request the 49-page paper, but a Corps spokesperson acknowledges receiving it.

Several scholars who track the convention say such activities undermine—if not openly violate—the chemical weapons treaty, which prohibits the development and use of chemical agents that cause “temporary incapacitation or permanent harm to humans or animals.” “This is definitely pushing the envelope, if not crossing the line, of what is covered in the treaty,” says Jonathan Tucker, who follows chemical and biological weapons for the Monterey Institute of International Studies' Washington, D.C., office. Tucker says the list of proscribed agents, which include nerve gas, mustard gas, and weapons containing commercial chemicals such as hydrogen cyanide and phosgene, is open-ended and based on the agents' ability to injure a target population.

The U.S. efforts also raise a red flag for Julian Perry-Robinson, a professor of science policy at the University of Sussex, U.K. Any work on such weapons, he says, “is historically troubling because it ties in to an older U.S. program” that disappeared from view during the Cold War. “It's worrying to see it coming up again.” New studies on nonlethal chemical agents, he adds, also send a message to other countries that it's all right to pursue research on more toxic agents.

U.S. military officials discussed nonlethal weapons at a joint U.S.-U.K. seminar held November 2000 at the Ministry of Defence in London. Pentagon officials there suggested that some of the military's research be funded by civilian agencies. According to an official report of the seminar, U.S. officials declared that “if there are promising technologies that the DOD is prohibited from pursuing,” the military should “set up MOA [memoranda of agreement] with DOJ and DOE [the Department of Energy].”

Some experts see the funding of Mazzara's work through DOJ as an example of this approach. “It's a pretty clear intent to violate the treaty,” says Tucker, “if the intent is to use these weapons in international military conflict.”

The National Academy of Sciences has just completed a review of the military's nonlethal weapons program, and Kenny's study was included in background material that the Marine Corps provided the panel. Negotiations over which portions of the report might be militarily sensitive have delayed its release, according to a spokesperson at the academy.

The Sunshine Project, a government watchdog group based in Austin, Texas, has publicized Kenny's study. And its director, Edward Hammond, is not waiting for the academy's verdict on the quality of the research: “It's shocking and disturbing that this kind of weapon would be contemplated at all,” he says.

16. METEOROLOGY

# Great Balls of Ice!

1. Xavier Bosch*
1. Xavier Bosch is a science writer in Barcelona.

Scientists are still struggling to explain a spate of large ice chunks that rained on Spain a couple of years ago

MADRID—After a football-sized chunk of ice plummeted from the sky on a sunny day in January 2000, smashing through the windshield of a parked car in Tocina, Jesús Martínez-Frías raced to the Andalusian village to retrieve the 2-kilogram object. Over the course of a week, the planetary geologist with the Madrid-based Center for Astrobiology gathered up several more projectiles. Then the phenomenon ended as suddenly as it had begun. But the mystery of where the ice balls came from lingered, and Spain's Higher Council of Scientific Research asked Martínez-Frías to lead a team to solve it.

Two and a half years later, the Spanish researchers have proposed a novel mechanism for generating “hail” on a clear day. It's an “important advance in that it thoroughly documents and provides an explanation for a spectacular phenomenon,” says geologist Roger Buick of the University of Washington, Seattle, who has studied an ice ball that fell from a clear sky in Australia last year. Other experts are far from convinced, in part because it would take so much time to accrete such large masses that the whole notion seems implausible. “I don't like to claim that anything is absolutely impossible, but this comes awfully close,” says Charles Knight, an expert on hail at the University Corporation for Atmospheric Research in Boulder, Colorado.

Like crop circles, ice balls could be the product of an ingenious hoax. However, the “amazing similarity in the details” of about 50 reported ice balls from all over the world over the past decade suggests that researchers have not been hoodwinked, insists David Travis, a climatologist at the University of Wisconsin, Whitewater. A few phony ice balls did turn up in Spain after the initial smattering, but the Spanish team says it was able to distinguish them from the ice balls that fell in January 2000 by their chemical and isotopic composition. The researchers focused on five of the original projectiles.

Martínez-Frías and his colleagues quickly dismissed other likely explanations. The ice balls could not have fallen from jets as frozen waste flushed from toilets because they lacked traces of urine or feces or the blue disinfectant used in airplane toilets. Nor were they likely to have been dislodged from a fuselage, the team maintains, as air traffic control records indicated that no planes overflew the region on the dates when two of the ice balls fell. They could not be of extraterrestrial origin—cometary debris, for instance—as they were made of rainwater.

That left researchers with the seeming conundrum of how to create massive chunks of ice in the absence of precipitation. Sawing into the ice balls provided clues. They were chock-full of air bubbles, had some onionlike layering, and contained gases such as ammonia and particulates such as silica, both of which get trapped in hailstones. The isotopic distribution of oxygen-18 and deuterium, as described in the June issue of the Journal of Chromatography A, also tracked that of hail.

To Martínez-Frías, the ice balls had hail written all over them. Hailstones form when ice crystals in a storm cloud collide with cloud water droplets at below-freezing temperatures. He and his colleagues suggest that a similar mechanism—ice crystals swept through cold, humid air pockets—could account for ice-ball formation. To support this notion, they cite a meteorological anomaly. Beginning on 7 January 2000—3 days before the first reported fall—NASA satellite imagery revealed that ozone levels were unusually low over southeastern Spain. In a well-known phenomenon, depressed ozone levels permit more solar radiation to reach the troposphere, triggering a concomitant cooling in the lower stratosphere. Around the time of the ice-ball falls, these temperature differences in the atmospheric layers created strong wind shears, Martínez-Frías says.

A second atypical phenomenon might also have been at play. A team led by meteorologist Millán Millán, head of the Mediterranean Center for Environmental Studies in Valencia, has found that the lower stratosphere was unusually moist at the time the ice balls fell. Millán estimates that a “condensation aggregate”—a growing ice ball—would free-fall about 19 kilometers through a nearly saturated atmosphere. The roughly 10 minutes of free-fall would be enough time for kilogram-sized ice balls to form, Millán argues: “We appear to be looking at nuclei that have descended through a very moist atmosphere, growing as they fell.” The source of the nuclei, Martínez-Frías and Travis suggest, could have been lingering jet contrails. Remnants of contrails can last for days as wispy cirrus clouds that are sometimes invisible to the naked eye. Many fewer nuclei would be available in these conditions than in a hail-forming thunderstorm, thus explaining why there were so few “megacryometeors,” as Martínez-Frías calls the ice balls.

Several experts contacted by Science reject the notion that hailstonelike processes can happen on clear days. “Solid ice cannot form in the absence of thick, highly visible clouds,” says Knight. He speculates that the conditions above Spain could have generated unusual precipitation—but not an ice ball. With “incredibly clean air and an incredibly long [free-fall] time,” he says, “the result would be an incredibly large snow crystal.” One atmospheric physicist suggests that an ice ball might form when rainwater pools in a cavity of a plane, then freezes and breaks off during flight. Commercial databases might not have logged military or private planes in the area when the ice balls fell, Knight says, adding, “the meteorological explanations just don't make sense to me.”

Martínez-Frías and his colleagues acknowledge that their mechanism is highly speculative. He is building an ice-ball library and forging links, through his Web site (tierra.rediris.es/megacryometeors), with scientists in other countries in pursuit of the mysterious objects. But it might take years to write the last chapter of this meteorological detective story—unless a fiendishly clever perpetrator confesses to an elaborate hoax.