# News this Week

Science  04 Jun 1999:
Vol. 284, Issue 5420, pp. 1594
1. SPALLATION SOURCE

# DOE Project Survives Close Call in Preliminary House Budget Vote

1. David Malakoff

Neutron scientists are breathing a bit easier after their flagship U.S. construction project, the Spallation Neutron Source (SNS), walked a political tightrope last week in Congress—and survived. The key vote was cast by the House Science Committee, which reversed a thumbs-down verdict earlier in the day and agreed on a $100 million budget to begin construction. That authorization, combined with$169 million approved the previous week by a Senate spending panel, puts the $1.36 billion accelerator on firmer footing for budget battles later in the year. The SNS, if built as planned at the Department of Energy's Oak Ridge (Tennessee) National Laboratory, will give scientists a uniquely powerful tool to probe the structure of matter, from proteins to metals, using a pulsed beam of neutrons (Science, 23 January 1998, p. 470). But as the largest new science project in the U.S. budget, it has drawn close attention from Congress. In particular, Representative James Sensenbrenner (R-WI), chair of the science committee, wants to make sure that the project's management structure, which teams five Department of Energy (DOE) national laboratories, won't produce a junior version of the failed$12 billion Superconducting Super Collider, an atom-smashing megaproject killed in midconstruction. As a result of that debacle, “the SNS is being held to a higher standard of review,” says Martha Krebs, head of DOE's Office of Science.

Indeed, reports earlier this year that the project was becoming hobbled by management troubles and delays led DOE to reshuffle its supervisory lineup. It recruited neutron scientist David Moncton from the Argonne National Laboratory in Illinois—where he successfully delivered the $812 million Advanced Photon Source on time and under budget—to lead the project. “We have recruited a manager whose record cannot be questioned,” crows DOE Under Secretary Ernest Moniz. However, Moncton's arrival wasn't enough to satisfy Sensenbrenner. In late March, after a visit to Oak Ridge, he announced his opposition to DOE's 2000 budget request for$196 million in construction funds until the project passed a July review and implemented other planned reforms, including giving Moncton greater authority over the project. At a 25 May committee meeting, he unveiled a bill that prohibited any SNS construction spending and challenged DOE to deliver on its promises before he would consider authorizing any money. “If DOE can get its act together, we can authorize this project,” he said.

But rather than wait to start the clock, Representative Jerry Costello (D-IL) proposed an amendment that would have restored $150 million in construction funds sooner. Sensenbrenner's bill “would effectively pull the plug on the nation's number one science project,” Costello charged, adding that the fiscal uncertainty would scare away recruits. He also warned that House appropriators—who actually set DOE's budget—would use Sensenbrenner's stance as an excuse to starve the project. After a heated debate, the amendment failed on a 17-17 vote and the committee recessed for lunch. During the break, Representative Bart Gordon (D-TN) drew up a compromise plan that included$100 million for construction once DOE met all of Sensenbrenner's conditions and offsetting cuts in other DOE programs. The amendment passed, 28-0. “At the end of the day, the committee had no reason to block construction,” Gordon said. “This is good news for maintaining the project's momentum,” Krebs declared.

Even so, the House vote virtually ensures that the SNS will receive less construction money next year than the $196 million requested. And that could pose a problem, say DOE officials, who argue that the project needs at least$150 million to stay on track. A final decision on the 2000 budget may not come until late this year.

2. ORGAN TRANSPLANTS

# New Drug Blocks Rejection in Monkeys

1. Gretchen Vogel

People with failing kidneys face a difficult choice: chronic dialysis or a kidney transplant with a lifetime of immunosuppressive drugs, which boost the risk of cancer and are themselves toxic to kidneys. But that choice may someday be easier, if tests of a new drug in monkeys eventually pan out in people.

In monkeys, blocking a key immune system signal for only a few months after a transplant leads to long-term acceptance of the new organ—with no detectable side effects, according to a report in the June issue of Nature Medicine by transplant immunologist Allan Kirk and endocrinologist David Harlan of the Naval Medical Research Center in Bethesda, Maryland, and their colleagues. Human trials are just getting under way, but the primate results are “really, truly amazing,” says transplant immunologist Norma Kenyon of the Diabetes Research Institute at the University of Miami in Florida.

The scientists caution that it's too soon to know if the monkeys have permanently accepted their new organs. The animals have developed antibodies to the transplanted kidneys, and although after more than a year those antibodies don't seem to be doing any harm, they may be the first signs of eventual rejection, says Kirk. But even with such caveats, “it is spectacular to have a monkey off of immunosuppression, with good graft function, for more than a year,” says transplant immunologist Christian Larsen of Emory University in Atlanta, Georgia.

The new drug is an antibody that binds to a protein called CD154, one of two signals that the immune system's T cells need to launch an attack against an invader. When T cells encounter a foreign molecule such as those on transplanted tissue, they become activated and produce more CD154, which in turn binds to a receptor called CD40 on other immune cells. That sends a signal for an all-out, devastating assault on the transplanted organ. The antibody is designed to block the attack by binding to CD154 and preventing it from binding to its receptor.

The strategy has so far surpassed expectations. Building on initial animal trials, the team transplanted new kidneys into 25 juvenile monkeys. To increase the challenge, donors and recipients were mismatched for the major histocompatibility complex, the series of proteins on cells that help the immune system distinguish between native and foreign cells. The mismatch had lethal consequences: Control monkeys given either no treatment or standard immunosuppressive drugs rejected the organs and died in 9 days or less.

But animals given the drug fared much better. A group of nine monkeys received weekly doses of the antibody for 1 month and monthly doses for 5 months afterward. All treatment was then stopped. More than a year after treatment, eight of that group are still alive and well. The one death was due to complications during a routine blood draw, and an autopsy revealed that the monkey had normally functioning kidneys when it died. Another monkey from a different trial received a transplant in March 1997, was given a month of therapy, and has been living off a mismatched kidney ever since. “It just works every time,” says Harlan of the drug. “I'm half expecting the balloon to pop at some point, but so far it hasn't.”

The technique, moreover, is not limited to kidney transplants. Kenyon's team has achieved similar success using the same antibody to block rejection of pancreas cell transplants in monkeys. In a paper in press at the Proceedings of the National Academy of Sciences, she reports that all six monkeys who received the CD154 antibody with their transplants have functioning islet cells as much as a year and a half later.

So far, Kirk and Harlan haven't been able to detect any side effects. Their monkeys have normal numbers of immune cells, did not develop wound infections, and responded normally to vaccines, indicating healthy immune function. Initial safety trials of the drug in humans, ongoing for more than a year, haven't produced any side effects, either. “I'm sure there are some downsides,” says Kirk, “but we haven't found them yet.”

One puzzling finding is that standard immunosuppressive drugs seem to interfere with the antibody's effect: Of the 11 animals that received a combination of standard immunosuppressors and the antibody, five died after acute rejection episodes. And no one is sure why the monkey's immune system accepts the foreign tissue long after the drug is stopped, although there are several theories.

Some scientists suspect that T cells activated by the foreign organ but lacking the CD154 signal simply die shortly after a transplant, although it's not clear why new generations of T cells wouldn't recognize and attack the tissue. Others suspect that the “stalled” T cells may somehow be protective. Biopsies of the transplanted kidneys reveal small clusters of activated T cells, which look like the ones seen in a few rare patients who, against doctors' orders, stop taking immunosuppressive drugs and surprisingly don't reject their transplants, says transplant immunologist Hans Sollinger of the University of Wisconsin, Madison. That may also help explain why the antibody performs poorly in combination with immunosuppressive drugs, which block T cells' initial activation.

That interference has made it difficult to design human trials for the new drug. Asking patients to forgo standard drugs to allow the experimental treatment to work is a “daunting” ethical challenge, says transplant immunologist Laurence Turka of the University of Pennsylvania. But the promising animal data prompted the National Institutes of Health last year to set aside a wing in its clinical center for trials of tolerance-inducing therapies, and researchers have enrolled a few patients for trials of a combination of the new drug and low doses of standard immunosuppression. If those prove as successful as the monkey trials, transplants may become an easier choice.

3. GERMAN SCIENCE POLICY

# Panel Calls for More Flexibility in Research

1. Robert Koenig

Germany's vaunted research system may be too rigid for its own good. Critics have accused it of being overly hierarchical and too slow to respond to hot research areas, and complained that it tends to hold back some young researchers by keeping them under the thumb of older professors. Last week, an international panel of prominent scientists echoed some of those gripes and went on to suggest a raft of reforms that aim to achieve more flexibility, greater cooperation between research institutes and universities, and give postdocs considerably more independence.

The report was requested by the Federal-State Commission for Educational Planning and Research Promotion (BLK), which represents research and higher education ministries at the state and national level. In a summary issued last week, the 10-member panel—led by materials scientist Richard Brook, chief executive of Britain's Engineering and Physical Sciences Research Council—praised the overall quality of German research, which is based on three pillars: universities, the Max Planck Society, and the DFG basic research granting agency. “This is a strong, high-achieving research system,” says Brook.

But the panel also highlights areas in need of improvement. It recommends that the traditionally independent Max Planck Society forge closer ties to universities and develop research groups that can respond more quickly to rapid new developments in science. In addition, it suggests that universities replace the post-Ph.D. “habilitation” qualification for aspiring professors with something like the U.S.-style “assistant professor” system, and that the DFG restructure its peer-review system and its strategy for promoting new disciplines. Overall, panelists found, the research system tends to be driven by middle managers, such as Max Planck institute directors. “There are great merits in a strong middle, but we'd also like to see a bit more life at the upper and lower levels,” says Brook, who directed a Max Planck institute in Stuttgart from 1988 to 1991.

After its yearlong inquiry, the panel concluded that closer cooperation between Max Planck institutes and university researchers might help improve what Brook calls the “mixed reputation” of German universities. Some scientists criticize Germany's university system for being too rigid, especially during the habilitation years. German federal research minister Edelgard Bulmahn, deputy chair of the BLK, has pointed out similar shortcomings. In a recent interview with Science, she said she wants to phase out the habilitation—a lengthy process during which postdocs do major projects under the strict supervision of professors—and bolster ties between university and nonuniversity research. Last week, Bulmahn called for “an intensive discussion” of the report's findings.

Brook says the panel found the DFG granting agency to have “a conservative nature” that could be revitalized by revamping aspects of its structure and programs, and perhaps by more actively steering researchers toward areas of research that it deems important. The report suggests a “strategically oriented program” for research grants, as well as a more active approach to funding progressive university programs, such as those supporting the early independence of young scientists. It also recommends opening up the DFG's peer-review system—for example, by including more women and younger researchers as reviewers.

Both the DFG and Max Planck responded swiftly to the report. The DFG's president, biochemist Ernst-Ludwig Winnacker, calls it “a thorough analysis” and says the DFG has already set up new funding programs for independent young scientists and is expanding its roster of peer reviewers. But Winnacker questions the suggestion that the DFG cherry-pick areas of new high-priority research: “The DFG cannot, must not, and should not compete with the federal and state governments, which are extensively involved in research funding that is guided by general political criteria.”

In a statement, Max Planck said it was already “well prepared” to implement some of the commission's suggestions, in part because the society is in the midst of an internal reassessment, and also because it has already taken steps to strengthen its connections to universities and to bolster its programs for young researchers. The society plans to establish several “International Max Planck Research Schools” near universities, increasing the number of Ph.D. students who conduct research at its institutes.

Brook says he expects German research to continue to thrive, especially if reforms are embraced: “It's much more difficult to evaluate a high-quality research system, such as Germany's, than a low-quality one.”

4. PHYSICS

# Come Fly With Me, Goldin Tells Physicists

1. James Glanz

BATAVIA, ILLINOIS—Space is the final frontier for particle physics, NASA Administrator Daniel Goldin declared in a 28 May press conference here at the Fermi National Accelerator Laboratory (Fermilab). But Goldin's vision of joining forces with the Department of Energy (DOE) and other agencies in an all-out assault on the mysteries of gravity and high-energy physics failed to uplift some listeners when he labeled Earth-bound accelerators—the focus of DOE's high-energy physics program—a “smokestack approach” to research.

The message of the press conference, which also included representatives from DOE and the National Science Foundation (NSF), was that the agencies are encouraging grant proposals submitted to them jointly. Goldin argued that because of the colossally high energies at play in the big bang, the neighborhood of black holes, and within neutron stars, these astrophysical phenomena should be regarded as physics experiments that dwarf anything that can be done on Earth. NASA satellites, he said, could exploit those natural experiments by collecting radiation or particles, insuring that physicists would no longer be “victims of the last [terrestrial] machine you built.”

Goldin's reference to scientists who focus on “the next bigger machine based on yesterday's technology,” however, ruffled some scientists at Fermilab, which was just four days away from dedicating its latest particle accelerator, the $260 million Main Injector. Physicist and DOE Under Secretary Ernest Moniz, who was seated next to Goldin, interjected politely that physicists can look forward to “important advances, as well, in accelerator-based experiments.” A subsequent speech by Goldin during a conference here had been purged of disparaging references to standard particle accelerators that appeared in earlier drafts and received generally positive reviews. “Dan Goldin's inspiring set of things to do was really spectacular, I thought,” said Leon Lederman, a Fermilab Nobel laureate in particle physics. “That's a grand vision,” added Scott Burles, an astronomer at the University of Chicago, although “you're going to have to be very clever to come up with [actual] missions.” Goldin was short on specifics, and Moniz, at the press conference, stipulated that no new advisory body would be formed to guide the effort, whose direction would instead be determined by individual proposals that passed peer review. Nor was there any mention of new money to fund the initiative. In his speech, however, Goldin said that several years ago, when he first proposed NASA's “Origins” program to study the origins of life in the cosmos, “we didn't have a nickel in the budget” for it, but funding materialized as his vision got fleshed out. In response to a question at the press conference, Goldin did show some interest in a concrete proposal by Saul Perlmutter of Lawrence Berkeley National Laboratory to build a telescopic satellite that would vastly expand both the quantity and precision of observations of distant supernovae, which have suggested that space is filled with a strange form of energy that counteracts gravity on large scales (Science, 18 December 1998, p. 2156). The satellite would rely on new charge-coupled device light-sensing technology developed at the DOE lab. “I think it's very exciting,” said David Spergel, a Princeton University cosmologist who is familiar with Perlmutter's concept. If NASA and DOE can get in the same flight pattern, particle physics may yet go where it has never gone before. 5. SYNTHETIC CHEMISTRY # 40 Steps to a Chemical Synthesis Summit 1. Robert F. Service Like mountaineers who set off to scale ever more challenging peaks, organic chemists over the past half-century have tested the limits of their skills by attempting to synthesize increasingly complicated natural molecules, such as antibiotics and steroid hormones. The most fiendishly complex targets have taken synthesis labs a decade or more to conquer. With the completion of each new project, labs scan the horizon for even higher peaks. And in the past 2 years, few mountaintops were more tantalizing than a pair of jellyfish-shaped molecules found in 1997 in a fungus. One enticement was the anticancer and cholesterol-lowering properties of the natural compounds, called CP molecules. The other was their complexity. The molecules' compact structure, crammed with chemical groups, made them “diabolical” targets, says K. C. Nicolaou, an organic chemist at The Scripps Research Institute in La Jolla and the University of California, San Diego. But in the 1 June issue of Angewandte Chemie, Nicolaou and his colleagues report having scaled that demonic peak: They have performed the first-ever complete synthesis of the CP molecules. “It's an extremely impressive accomplishment,” says Samuel Danishefsky, whose own group at Columbia University in New York City was closing in on the same goal. Other recently synthesized molecules have been more than four times the size of the CPs, which have 31 carbon atoms each. But Danishefsky says the CP molecules require particular finesse. “The functional groups bump into each other so that it's difficult to work on one portion of the molecule without affecting another part,” he says. The CP molecules originally attracted attention when researchers at the pharmaceutical giant Pfizer showed that they inhibited the work of a cancer-causing gene known as Ras, which is overactive in up to 80% of human cancers. CPs, it turns out, block the addition of a chemical group known as a farnesyl group onto the Ras gene, a key step in its activation. Other more potent farnesyl blockers have been discovered, says Takushi Kaneko, a medicinal chemist at Pfizer's research center in Groton, Connecticut, who helped nail down the CP molecules' structure. But the new synthesis work could still prove vital, he says, by allowing chemists to manufacture CP analogs that may prove even more potent and also easier to produce than the CPs themselves. Getting this far was a nearly 2-year slog. In all, it took more than 40 chemical steps and many grams of starting materials to make milligrams of the molecules, which consist of a core ring of nine carbon atoms bearing three more carbon-oxygen rings. And twice the group had progressed to key intermediate compounds along the way, only to find that although they were only a few bonds away from the complete structure, they could not forge the final links. The final attempt that got them to the summit took three key steps. First, the researchers had to convert a linear hydrocarbon precursor molecule into the nine-membered ring at the core of each CP molecule. They turned to a well-known ring-forming process known as an intramolecular Diels-Alder reaction and tweaked the reaction conditions to coax the precursor to adopt the correct ring-shaped structure. For the next step, the Scripps researchers developed a set of novel “cascade” reactions. Cascade reactions run through a staccato of intermediate steps—each one automatically producing the right materials and conditions for the next—before ending up at a final product. The researchers used two of their cascade reactions to fuse two additional five-membered carbon and oxygen rings to opposite sides of the core. A final summit push, consisting of a flurry of reactions, provided them with one of the CPs, called CP-263,114, the more stable of the pair. But they also wanted to make its partner, CP-225,917, which differs only in that one of the three attached rings is broken, the frayed ends capped with hydroxyl groups. Trying to coax the stable CP-263,114 into a more unstable form proved very difficult. After numerous attempts, the team designed another cascade reaction, which finished the job. When the resulting compound passed muster in a structure-determining nuclear magnetic resonance machine, the climb was complete. Atop the mountain, says Scripps Ph.D. student Phil Baran, “it feels like a 200-ton anvil has lifted off my back.” Yet in some ways the work is just beginning. Now the hunt is on to come up with CP analogs that are more potent and simpler to make. The Scripps team is also launching studies of the detailed biological effects of CPs and their kin. Of course, the search is also on for new molecular mountains to climb. 6. BEHAVIORAL GENETICS # Fickle Mice Highlight Test Problems 1. Martin Enserink Studying the genetics of behavior is often like riding a roller coaster. A standard way to look for the genetic basis of a behavior—anxiety, say, or aggression—is to knock out a suspect gene in a mouse strain and test the animals in the laboratory. But no sooner has one group of researchers tied a gene to a behavior when along comes the next study, proving that the link is spurious or even that the gene in question has exactly the opposite effect. Now, on page 1670, a study born in part out of frustration over this phenomenon shows how easily such discrepancies may arise. Behavioral geneticists from three labs across North America applied the same battery of behavioral tests to the same strains of mice, under almost exactly the same circumstances—and yet they often got strikingly different results. This implies that almost undetectable environmental differences may have large behavioral consequences. The finding is bound to complicate efforts to pin down the genetic influences on behavior. “It's the kind of study that needs to be done, but nobody wants to be doing,” says behavioral neuroscientist Elizabeth Simpson of the University of British Columbia in Vancouver. “You're looking into something that people would like to believe is not a problem.” The three labs, led by John Crabbe, a behavioral geneticist at the Veterans Affairs' Portland Alcohol Research Center and Oregon Health Sciences University, Douglas Wahlsten of the University of Alberta in Edmonton, Canada, and Bruce Dudek of the State University of New York, Albany, carefully standardized the tests. All three started on 20 April 1998 between 8:30 and 9:00 a.m. local time, and each used a total of 128 77-day-old mice from the same eight strains. Conditions in the three labs, from the light-dark cycle to the brand of mouse feed, were painstakingly equated, to the point of driving the researchers “nuts,” says Crabbe. And yet, genetically identical mice often behaved differently, depending on where they were tested. One puzzling result came from a standard test for anxiety, the so-called “elevated plus-maze.” In this test, researchers place a mouse in the center of a big horizontal plus sign fixed about 1 meter above the floor and then measure how much time it spends in each of the four arms, two of which have transparent plastic walls, while the other two are open. Animals that prefer the safety of the walled arms are thought to be anxious, while the ones that venture onto the open arms, nosily peering over the cliffs, are deemed less inhibited. As it turned out, anxiety levels among mice of all strains were lowest in Edmonton. In addition, one strain, in which a receptor for the neurotransmitter molecule serotonin was knocked out, gave different results in all three cities: In Portland, it showed more activity on the maze than controls with intact serotonin receptors; in Albany, it was less active; and in Edmonton, lacking the receptor didn't seem to make any difference. The same mutant also provided an unpleasant shock for Crabbe. In 1996, his team reported in Nature Genetics that the animals drank much more alcohol than control mice having the receptor—a major result, addiction researchers said, because it seemed to firmly nail the importance of the serotonin pathway in addiction. The team had replicated the finding four times. But this time, all three teams found that the animals were no fonder of drink than controls. “It was a bad surprise,” says Crabbe, who is now trying to find subtle genetic changes that may have caused the mutants to lose their taste for alcohol. As for the other discrepancies, the researchers can only conclude, Crabbe says, that they are the result of very subtle differences in lab conditions, like the chemical composition of the water, or the way the researchers handled the animals, or even the way the scientists and technicians looked or smelled. In Edmonton, a research assistant was highly allergic to mice and wore a respirator. “That looks weird to us; it may look strange to a mouse, too,” says Crabbe. Crabbe doesn't have a clear solution yet to the problems that the study lays bare, but he is now planning experiments to find out if a combination of three or more different tests, designed to measure the same behavioral trait in different ways, would produce more reproducible results than a single test. Meanwhile, he says, the field of behavioral genetics should at least standardize its tests and perform them just as attentively as, say, a DNA extraction. But because even that won't eliminate outcome differences, every result should be replicated with a new batch of mice within the same lab, and perhaps even elsewhere, before it's published. 7. EXPERT TESTIMONY # Project Offers Judges Neutral Science Advice 1. Jocelyn Kaiser Federal judges looking for impartial scientists to help sift through complex technical evidence will soon have an easy way to find them. Last week the American Association for the Advancement of Science (AAAS, publisher of Science) launched a 5-year pilot project to supply judges with lists of experts who can provide advice in complicated cases, such as claims of software patent infringement or illness from exposure to a toxic substance. The project is intended to cut through the legal confusion generated when expert witnesses hired by each side dispute the significance of such evidence. Pamela Ann Rymer, a judge with the U.S. Court of Appeals for the Ninth Circuit in Pasadena, California, and chair of the project's advisory committee, welcomes the arrival, “for the first time, [of] a single, independent, and neutral source for identifying potential experts.” The$500,000 project, with support from the Leland Fikes Foundation of Dallas and financier George Soros's Open Society Institute, grows out of a 9-year-old idea from a joint panel of AAAS and the American Bar Association. Its stock rose after a 1993 Supreme Court decision urged judges to act as “gatekeepers” for scientific evidence, disallowing expert testimony that doesn't meet accepted scientific standards. The court expanded the ruling earlier this year to include evidence involving the expertise of engineers and those in other technical fields (Science, 2 April, p. 21).

The Court Appointed Scientific Experts (CASE) project will rely on a panel of scientists, professional societies, and even the Internet “to help identify people respected in their field” in response to a request from a judge, says project manager Deborah Runkle. It will also develop guidelines for measuring potential conflicts of interest. A 1993 survey by the Federal Judicial Center (FJC) suggests that district judges would welcome input from an impartial expert, who might be involved in everything from tutoring judges and juries in computer codes to explaining the methodology underlying scientific testimony.

Why would a scientist want to get involved in a court case? “It's a service,” says Runkle, just like testifying before Congress or serving on a blue-ribbon panel. Adds CASE advisory panel member Sheila Widnall, an aeronautics professor at the Massachusetts Institute of Technology, “It's extremely important for our society, as issues get more and more complicated, that there is a voice” from scientists. The FJC will evaluate the project's impact after 5 years.

Opening for business this fall, the project has already drawn criticism from trial lawyers, who fear that it could tip the balance in the current adversarial system. Ned Miltenberg of the Association of Trial Lawyers of America in Washington, D.C. says that scientists are deeply divided on such legal hot buttons as whether animal tests should be admitted as evidence of causation in toxic tort cases. Putting a stamp of approval on a witness, says Miltenberg, could easily lead “overawed” judges and juries to defer to the court-appointed expert's opinion.

But Runkle says “any good scientist” will acknowledge legitimate opposing views. Besides, she adds, conflicts “are going to exist whether we are here or not.”

8. CELL BIOLOGY

# New Leads to Cancer, Arthritis Therapies

1. Michael Hagmann

Some enzymes get the glamorous jobs: repairing damaged DNA, for example, or shuttling other substances into and out of cells. Others pursue seemingly boring occupations. The enzymes known as metalloproteinases, for instance, simply chew up proteins. They are every bit as essential, however. By breaking down collagen and the other proteins that make up connective tissue, they remodel the entire body during embryonic development and help migrating cells, such as immune cells or the cells necessary for wound healing, move to where they are needed. And like their glamorous cousins, the protein-degrading molecules can cause serious problems when they become overactive, allowing cancers to spread or eroding joints in arthritis. Now, two new findings about the metalloproteinases could open the way to controlling the enzymes.

On page 1667, Karl Tryggvason and his colleagues at the Karolinska Institute in Stockholm, Sweden, describe for the first time the complete three-dimensional structure of a metalloproteinase. This enzyme, known as MMP-2 (for matrix metalloproteinase 2), is usually found only in the developing embryo and healing wounds. But it can also help cancer cells spread in the body and allow growing tumors to build new blood supply lines, and so researchers have been looking for drugs that inhibit its action—a quest that the new structure may aid.

And on page 1664, a team led by Elizabeth Arner at DuPont Pharmaceuticals Co. in Wilmington, Delaware, reports the cloning of a new metalloproteinase that seems to play a key role in the development of arthritis by breaking down a cartilage protein called aggrecan; it may thus be a target for antiarthritis drugs. “Everyone looking for arthritis targets is very excited, [because] understanding cartilage destruction had been on hold until the ‘aggrecanase’ activity was found. This will set the stage for a lot of activity,” predicts arthritis expert John Sandy of the Shriners Hospital for Children in Tampa, Florida.

The MMPs, so named for the zinc ion in their catalytic centers, first appeared in the late 1960s when MMP-1, a collagen-degrading enzyme, was discovered. After a long gap, in 1980, Lance Liotta of the National Cancer Institute (NCI), working with Tryggvason, found that various tumor cell lines produce huge amounts of a related enzyme, MMP-2, and linked the enzyme to metastasis. “MMP-2 is not active in benign tumors; it only becomes activated once the tumors become invasive,” Tryggvason explains. The enzyme apparently helps tumor cells spread by degrading the collagen in basement membranes, a protein mesh around blood vessels and other organs that is usually the first barrier roaming tumor cells have to cross.

Now Tryggvason has produced the first picture of this enzyme in its inactive form. Because perpetually active MMPs would be dangerous to the body—“they can literally dissolve you, so you don't want them floating around freely,” Tryggvason says—the enzymes are secreted as inactive proenzymes with one end of the molecule, the propeptide, serving as a built-in inhibitor that has to be cut off before the enzyme can act. Tryggvason's x-ray crystallographic structure of this molecule reveals several regions that might be targets for anticancer drugs.

One of these is, of course, the enzyme's active site—the part of the molecule that actually clips the substrate proteins. But previous structural studies of just the catalytic domains of several other MMPs showed that they are very similar, and this one is no exception. This similarity could make it difficult to design a specific inhibitor for MMP-2 that works by blocking the enzyme's active site.

But the proMMP-2 structure does reveal more promising targets. The inhibitory propeptide folds into several staircaselike helices, one of which completely shields the catalytic center. To activate the molecule, other proteins cleave a loop connecting the helices. This disrupts the structure of the propeptide leading to its release from the active site. The loop region in the propeptide could serve as a model for designing inhibitory drugs. “One could think of generating peptide analogs [similar to the loop region] that could block the activation step” by choking the MMP-2-activating enzyme, says William Stetler-Stevenson of the NCI's Extracellular Matrix Pathology section. Another MMP-2 region, the part that targets it to its substrate and has now been shown to have a clover leaf-like structure, might also be a drug target, especially as it has been found only in MMP-2 and another cancer villain, MMP-9.

The hunt will also be on for inhibitors of the aggrecanase enzyme, which could lead to arthritis treatments. Researchers have known for more than 20 years that the destruction of aggrecan is one of the early hallmarks of arthritis. But the original suspects, the MMPs, eventually proved innocent when researchers found that they cut aggrecan in the wrong places and failed to produce the characteristic pieces found in the synovial fluid—the lubricant—of arthritic joints. Hence, they postulated that an enigmatic “aggrecanase” was at fault.

That's pretty much how things stood until DuPont's Arner decided to take what might be called a brute-force approach, hoping that a sensitive screen applied to a large sample of cartilage would snare the molecule. Her team chopped up some 30 cow noses—“a good source for cartilage,” she says—and cultured the pieces for several weeks with interleukin-1, an inflammatory protein known to bolster cartilage breakdown by inducing aggrecanase production in the cartilage cells. She and her colleagues then ran their 30-plus liters of culture fluid through several purification steps, the last of which used a metalloproteinase inhibitor as bait. Before they did that, however, they saturated the culture fluid with another inhibitor that specifically binds the MMPs, so that those enzymes could not take the bait, while any aggrecanase present could. And indeed, her team ended up with a minute 10 micrograms of a single protein that showed the arthritis-specific aggrecan cleavage signature.

Next, they determined partial amino acid sequences of the protein, translated them back into the language of DNA, and searched DNA databases. They came up with a single hit, a portion of a mouse protein of unknown function that is more than 90% identical. Using this sequence information, the researchers then fished the corresponding human gene from a heart DNA library.

Aggrecanase-1, it turns out, is also a metalloproteinase, but it belongs to the so-called ADAMTS family that is only distantly related to the MMPs. The big question now is whether the enzyme is the main culprit in arthritis. Using a similar purification procedure, Arner and her colleagues have fished out another enzyme, aggrecanase-2, with properties very similar to those of aggrecanase-1. “Whether [aggrecanase-1] is the only player in arthritis is still out there in the open,” says cell biologist Judith White of the University of Virginia, Charlottesville. The only way to tell for sure is to inactivate the aggrecanase-1 and −2 genes and see whether that prevents or retards development of arthritis in mice.

But even if it does, drug developers will have to deal with a specificity problem similar to that with MMP-2. Not only are there at least 11 ADAMTS gene sequences reported thus far, but to make matters worse, aggrecanase-1 is not exclusive to joint cartilage. Arner's team found it in the heart, lungs, and brain as well. What it does there is not yet known, but the finding suggests that inhibiting the enzyme might cause unwanted side effects.

Still, as Amanda Fosang of the University of Melbourne, Australia, points out, “This is only the beginning. There's a lot more to come.” Arner is quick to agree. “I think I'm going to be pretty busy in the next few years.”

9. UNIVERSITY FUNDING

# Japan Wants Results to Influence Budgets

1. Dennis Normile

TOKYO—Japan's Education Ministry is weighing a plan to make an institution's track record a criterion in future spending on new research buildings and large equipment. The approach would break with the current practice of allocating infrastructure funds to universities and the ministry's institutes through a formula based primarily on size and tradition. Most scientists applaud the idea of a more rigorous evaluation of research programs, although some are concerned about what measures would be used and how the process might affect the research enterprise. “Outside evaluations are necessary,” says Kozi Nakai, a physicist at the private Science University of Tokyo. But he warns that, if the evaluations are not done right, “they could be dangerous.”

The recommendations were contained in an interim report released last week by the Science and Technology Council, which advises the Ministry of Education, Science, Sports, and Culture (Monbusho) on research matters. The report echoes a similar call last November by the University Council, which reviews ministry policy on higher education, to carry out evaluations of all educational programs. Evaluations of research efforts are likely to be based on both objective data, such as the numbers of recent papers and citations by faculty, and onsite program reviews by expert committees. Details will be worked out over the next year and made part of the mandate of a new evaluation body under Monbusho.

The new policy, if adopted, would mark a significant step in an ongoing campaign to increase the level of accountability at Japan's universities and research institutes. Until fairly recently, universities and professors operated free of almost any type of oversight. Faculty members arrived on campus with jobs for life and received guaranteed, although modest, funding for research even if they never published a paper or submitted a patent application. Infrastructure funds for the 98 national universities were split depending on enrollment and tradition and rarely varied.

That system has started to crumble in the past decade as the amount of research money made available to scientists through competitive, peer-reviewed grant schemes has nearly tripled, to $1.9 billion, while the base funding has grown more slowly, to$1.3 billion. “Individual researchers are now largely subjected to peer review of their work,” says Masayuki Shibata, director of Monbusho's Office of Science Policy.

Progress has been less dramatic on the institutional level, however. Although some departments invite blue-ribbon panels from around the world to review their research and education programs, most universities have opted for committees drawn from within. “These self-assessments just don't go far enough in an era when science and technology have become borderless,” says Hiroyuki Abe, president of Tohoku University in Sendai and chair of the science council's working group on research evaluation. The council's report recommends “third party,” or external, evaluations to remedy that problem.

Although the use of outside evaluations to shape government funding decisions for institutions has broad support, some scientists are worried about the details. Nakai, who also served on the evaluation working group, fears a centralized evaluation would tend to apply the same evaluation criteria across disciplines and between fields. The criteria for groups working in accelerator physics should be different from other areas of physics, not to mention engineering, he says, noting the variations between large and small science and the differing attitudes among disciplines toward the importance of publication. He also worries that a centralized system will seek a common denominator and penalize universities that cater to local needs. “Everyone recognizes that Japanese primary and secondary education is overly focused on the single objective of doing well on university entrance exams,” he says. “There is a possibility of making the same sort of mistake on this.”

Abe, who also serves on a new committee that will advise Monbusho on the envisioned evaluation body, says that policy-makers and advisers share these concerns. The report, which is intended to “advance the nation's scientific research,” also recommends that Monbusho create a mechanism that allows universities to collect overhead on government-sponsored research and improve ties between universities and the private sector. The report also asks Monbusho to revise rules that restrict the hiring of research assistants and technicians.

10. ASTRONOMY

# The Coolest Brown Dwarfs Proliferate

1. James Glanz

CHICAGO—By collecting and cataloging hundreds of millions of celestial objects, the Sloan Digital Sky Survey may turn a rare oddball into a common denizen of the heavens. The $80 million project is designed to create a three-dimensional map of galaxies extending to hundreds of millions of light-years. But the Sloan, which is still in a shakedown phase, has demonstrated that its census-taking power extends to our cosmic neighborhood as well. At an American Astronomical Society meeting here on Monday, survey members announced that it has turned up two of the coolest, dimmest stars called brown dwarfs ever seen, lurking by themselves in the equatorial sky. Probably located within 30 or 40 light-years of Earth, the brown dwarfs are so small that their surface temperature is no more than 1000 Kelvin, cool enough for methane and water—compounds normally associated with planets—to survive. Until now these molecules had been seen in only one other brown dwarf (Science, 1 December 1995, p. 1435), which was tethered to a brighter and more massive companion. “They're the first methane brown dwarfs found floating out there in isolation,” says David Golimowski of Johns Hopkins University and the Sloan team—and a hint that, as theorists have predicted, large numbers of such stars are waiting to be found. “Given that the sky survey was built for other purposes, that's a really handsome payoff,” says Alan Boss, an astrophysicist at the Carnegie Institution of Washington. The Sloan's special wide-field telescope on Apache Peak, New Mexico, is combing wide swaths of sky to compile a “field guide” of hundreds of millions of objects, from which the million brightest galaxies can be selected for detailed mapping. Like a trawl of the heavens, the effort also nets objects ranging from nearby stars to the most distant quasars—brilliant beacons in the early universe (Science, 11 December 1998, p. 1969). Sloan astronomers stumbled across the brown dwarfs while searching for new quasars in data from a test period. Because quasars are so distant that their light is shifted toward the red end of the spectrum, they glimmer in the Sloan's long-wavelength, infrared channels but vanish in optical bands. So do brown dwarfs, as Michael Strauss and Xiaohui Fan of Princeton University and Zlatan Tsvetanov and Wei Zheng of Johns Hopkins found when detailed studies of the objects' spectra revealed that they were cool, dim, and close by. Defined as stars with less than about 8% of the mass of the sun (or 80 Jupiter masses), brown dwarfs never ignite much fusion burning in their cores and gradually dim after they form. Dozens of garden-variety brown dwarfs, called L dwarfs, slightly too massive and warm to host water and methane, have turned up over the last couple of years, many of them in the infrared 2-Micron All-Sky Survey at the University of Massachusetts, Amherst. But theorists expected that the smaller methane dwarfs would be common as well, because a cloud of gas as small as 10 Jupiter masses can, in principle, collapse into a star. “It's nice to see some confirmation of that theoretical prejudice” from the Sloan results, says Boss. As the Sloan expands its view to cover much of the northern sky, the smallest dwarfs could become a tribe. “They're oddballs at the moment,” says Golimowski, “but I'm confident they'll be pretty boring objects within a few months.” 11. SCIENCE AND SOCIETY # European Researchers Grapple With Animal Rights 1. Robert Koenig Following a peak in animal rights extremism last year, researchers are now working to defuse the confrontation through concessions, better PR, and bringing the activists to the meeting table BREMEN—In a building surrounded by a chain-link fence topped with razor wire, neuroscientist Andreas Kreiter conducts his experiments in an atmosphere more akin to that of a top-secret weapons lab than a university research institute. The reason for the security: Kreiter and his laboratory have been the target of attacks and threats by animal rights activists for more than 2 years. A 3-meter-wide poster set up in downtown Bremen in 1997 branded him a “monkey torturer” and listed his home and lab addresses. Both he and his family have been subjected to death threats, and armed police escort him when he makes public appearances. Across the North Sea, in the normally peaceful atmosphere of Oxford University, physiologist Colin Blakemore—who uses laboratory animals to research brain development and cognition—has been beaten up by animal rights guerrillas and has been the target of letter bombs mailed to his home. Blakemore has been forced to install high fences, electronic detectors, and other security measures at his home to protect himself and his family. He says that defending himself and his work from such threats “has taken 3 or 4 years out of my research” since he was first threatened a dozen years ago. As the number one targets for activists in Germany and Britain, Kreiter and Blakemore are at the sharp end of European researchers' struggle against animal rights extremism. They are far from alone. Dozens of other scientists in northern Europe who are outspoken in defending their animal research have found themselves slandered on Web sites, attacked by demonstrators, or have had their labs damaged or disrupted by guerrilla assaults. “Some of these groups are vitriolic, and you have to worry about the threats,” says Denis Duboule, a developmental biologist at the University of Geneva who uses mice in his research. When he received an award for his work last year, activists tried to disrupt the ceremony, demonstrated at Duboule's lab, and spray-painted a warning on his home. The United Kingdom and Germany have some of the strictest animal protection regulations in the world and have long had active animal rights movements whose militancy has waxed and waned over past decades. In contrast to the United States, where animal rights activism has been in a lull in recent years, extremism surged in both countries last fall. In Britain, militant activist Barry Horne—in jail for firebombing shops he felt had supported animal experimentation—held a hunger strike in a widely publicized effort to win concessions from the government. British newspapers reported assassination threats by militant groups against researchers, including Blakemore, and others if Horne died. (He called off his hunger strike when he was near death.) In Germany, Kreiter and his research mentor—neuroscientist Wolf Singer, a director of the Max Planck Institute for Brain Research in Frankfurt—were also the targets of highly publicized threats. Max Planck president Hubert Markl called such intimidation “an attack on the freedom of research in Germany.” This heightened tension turned animal protection into a hot political issue. It also spurred efforts to blunt the tactics of the extremists and find common ground with moderate animal welfare advocates. Across the continent, bodies such as the European Union, the European Science Foundation, and Britain's Royal Society have commissioned reports or convened meetings to study and debate the issue. A British initiative to bring researchers and animal rights activists together to hammer out their differences may be followed in other countries. And in Germany, some mainstream political parties are backing a constitutional amendment that would guarantee animal rights—a move that is causing consternation among many researchers. The tensions have abated recently, perhaps in part because of these developments. Britain's Scotland Yard set up a special police unit to investigate violent animal rights terrorists, and police crackdowns in both countries have succeeded in arresting some of the most violent activists. “The animal rights militants have been shifting lately to more focused demonstrations, targeting specific places, such as the breeders of laboratory animals,” says Mark Matfield, who heads Britain's Research Defence Society and is also director of the European Biomedical Research Association, which represents similar groups that support researchers in 22 countries. In the most recent attacks, militants have targeted Britain's only commercial breeder of cats for research, Hillgrove Farm. “But these are just changes in tactics, not in the activists' ultimate goal of eliminating the use of animals in research,” Matfield adds. As one scientist says, “we are in the calm before the next storm.” ## War of words The surge in animal rights militancy last year provoked a wide range of responses among both researchers and European governments. One reaction was the urge to fight back—in the war of words at least. The animal rights movement is winning public sympathy, researchers argue, by distorting the truth about animal research and ignoring the benefits it brings. Ernst-Ludwig Winnacker, president of Germany's DFG basic research granting agency, says that about 80% of Germans surveyed in a recent poll said they oppose the use of animals for lab research; but when asked more specifically, a solid majority said they support such experimentation if it provides benefits for human medicine. A poll last month commissioned by Britain's New Scientist magazine found that 64% of 2009 people surveyed initially said they disagreed with animal experimentation, but when shown how such research would likely hasten progress in medicine, a slim majority favored such research. In early May, Ivar Aune, director of Germany's main lobbying group on behalf of researchers, the Gesellschaft Gesundheit und Forschung, told the Scandinavian Society for Laboratory Animal Science that researchers and the groups that represent them must take a more proactive approach to countering animal rights activism, by responding to inaccurate claims, providing accurate information to the public, and making their views known to key decision- makers. In May, the Royal Society's animal experiments committee met to discuss possible efforts to explain the position of science on the issue to the general public. A second response was to try to build bridges between researchers and animal activists. The British Home Office, which regulates animal experimentation, announced last month that it will sponsor a major forum in London on 9 July devoted to science and animal welfare—with participants including representatives from many of Britain's main research organizations and animal welfare groups. A special unit created a few years ago by the European Union—the European Centre for the Validation of Alternative Methods (ECVAM)—is working to promote dialogue among legislators, industries, biomedical scientists, consumer organizations, and animal welfare groups, to help develop and validate alternative test methods for research that would reduce, refine, or replace the use of lab animals. “The prospects for making steady progress are very good,” says ECVAM's director, Michael Balls, but he adds that “many individuals, especially in government and in animal welfare, have unrealistic expectations of the rate at which progress can be made in replacing current animal procedures.” But perhaps the most ambitious attempt to defuse tensions between animal rights groups and researchers is the Boyd Group. This initiative, begun in 1992 by Blakemore and Scottish animal welfare activist Les Ward, director of the moderate Advocates for Animals campaign, convenes meetings of representatives from both sides of the debate—such as the Royal Medical Colleges, the Wellcome Trust, the Medical Research Council, the British Association for the Advancement of Science, mainstream animal welfare and animal rights organizations, medical ethicists, and philosophers—to discuss issues and try to find some common ground. “We have tried to build some bridges,” says Ward, who advocates the eventual ending of all experimentation with animals. “This is a forum where both sides—rather than slugging it out in the media—can sit around a table and find some issues on which we can agree to reduce the pain and suffering of animals.” Although the group has reached general agreement in the past on issues such as animal experimentation for new cosmetic ingredients, a working group—which includes Ward and Blakemore—is now trying to find common ground on the tougher issue of primate research. “The Boyd Group is important,” says Blakemore, who would like to draw in representatives of more militant animal rights groups. “If you come face to face with a rational person and sit down and discuss a contentious issue, that defuses the potential violence and elevates the quality of debate.” An effort is also under way to establish a continental group—the “European Consensus Platform on Alternative Methods”—that would include representatives of government ministries, universities, animal welfare groups, and industries that use animals for research (mainly pharmaceutical, chemical, and cosmetics firms). A leader of that effort, Bernward Garthoff of Germany's Bayer AG drug and agrochemical firm, says he expects the new organization to begin with existing groups in Germany, Belgium, and the Netherlands. But it may expand later to include other nations, and it is applying for a “networking” grant under the European Union's Framework 5 Program. One goal of the platform group would be to identify further possible changes in global legal requirements for industrial-product testing that would decrease the number of animals required for experiments. ## Meeting halfway Yet another approach is to make concessions to the animal rights movement to try to draw support away from the more violent militants. Soon after its election 2 years ago, the British Labour government made some symbolic moves to assuage animal welfare groups, such as banning the use of animals to test new ingredients for cosmetics and for alcohol and tobacco products. But, Blakemore says, the government “has not really drawn the line” in defense of animal use for “the core areas of biomedical research where it is essential.” Politicians in Germany, however, are now considering a more risky strategy, which is stirring up the research community. Supported by animal welfare groups, an array of politicians is backing a constitutional amendment that would guarantee the welfare of animals. Although the wording of the proposed amendment is bland, Kreiter and other researchers contend that it could give animal rights groups the power to force many German researchers to go to court to gain approval for their projects. “This would make it more difficult to perform some experiments, because animal-protection groups have said publicly that they would take legal action against researchers,” says Winnacker of the DFG, which issued a statement this year opposing the amendment. Even some German scientists who are active in protecting animals in research labs warn that a constitutional amendment would be harmful to research. Hansjoachim Hackbarth, a professor of animal welfare and behavior at the University of Hannover's veterinary school, told Science that the proposed constitutional change “would stop or delay a great deal of research in Germany for a long time”—at least until court challenges are sorted out. “I oppose this, but I expect the German parliament to approve the amendment later this year.” Indeed, that prospect is looking increasingly possible. In April, the German cabinet—a Social Democrat and Green Party coalition—endorsed the concept of a constitutional amendment, which is now being considered by a parliamentary committee. In the past, the main barrier to parliamentary approval of such an amendment has been the center-right Christian Democratic Union (CDU), which could block the amendment from gaining the required two-thirds majority in parliament. But former German research minister Jürgen Rüttgers and some other CDU leaders now support a watered-down version of the amendment, which scientists say might still lead to court challenges against research. Despite efforts to tackle the problem, many biomedical scientists view with alarm the tightening noose of government restrictions and concessions made to the animal rights movement. “In Germany, it now seems to be easier to get permission for [an experimental] study involving humans than it is to get permission for animal experiments,” says Kreiter. “The trend frightens me. I see a tendency to try to balance out animals against humans.” He and Blakemore urge their fellow scientists to speak out forcefully against undue restrictions on research. Says Blakemore: “Somebody has to fight this fight, or the victory goes by default to the opposition. And that would have a terrible impact on science and medical care.” 12. SCIENCE AND SOCIETY # Doing Research Under Siege 1. Robert Koenig BREMEN—Andreas Kreiter is puzzled by the notoriety that now surrounds his work. He was quite open about his research plans when the University of Bremen hired him in 1996, and the brain research he now carries out on macaques has not only been through the mill of Germany's strict animal experimentation regulations, but it also uses standard procedures practiced by neuroscientists elsewhere. “We try to keep them active,” he says, pointing to four young macaques clambering over rope nets, adding that healthy animals are essential for his research. Kreiter, who heads a research team at the university's Center for Cognitive Neuroscience, studies the synchronization between nerve cells that are important in coding information in the brain. In his experiments with macaques—standard subjects for this kind of research—scientists surgically place an implant on the monkeys' skulls, using a small cylindrical tube through which hair-thin electrodes are inserted into the brain. Neural activity is measured while the monkeys carry out certain tasks. Kreiter says that the implants may look odd, but they “do not hurt or bother the monkeys,” and inserting the electrodes “is a painless procedure; the brain has no pain receptors.” He notes that “there are many U.S. scientists working with the same species of monkeys, and the techniques used for this type of brain research are almost identical everywhere.” All was normal in his lab, Kreiter says, until April 1997, when there was a sudden “explosion of opposition.” Animal rights activists threw red paint at a lab that they thought was his and set up a defamatory poster of Kreiter on a downtown street, listing his home and lab addresses. Six months later, he needed police protection, because demonstrators were lined up outside his office. Since then, Kreiter has been the prime target of the German animal rights movement, in part, he believes, simply because the head of Germany's biggest animal welfare group (the Deutsche Tierschutzbund) also lives in Bremen. “They spread horror tales in the local newspapers,” says Kreiter. “It steals a lot of my research time, and the controversy preoccupies me. … In the first few months, I would be worrying all the time: Is my house still standing? Are my children OK?” Soon it was not just the animal activists he had to worry about. Last fall, about 100 faculty members at Bremen University signed a petition opposing Kreiter's research. This was swiftly followed by a counterpetition signed by dozens of other Bremen professors supporting him. As the attacks and threats continued, the wider neuroscience community became involved. Earlier this year, nearly 50 U.S. and other international neuroscientists—including two Nobel laureates—signed a petition supporting Kreiter, calling him “a respected investigator in the field of systems and cognitive neuroscience. … The use of alert, behaving monkeys, which is the experimental preparation at issue in Professor Kreiter's case, is one of the best approaches currently at our disposal for addressing questions about the biological basis of higher brain functions.” The controversy has made it harder for him to carry out his research. “This makes it more difficult to hire talented young scientists. If one has a choice between a lab that is under attack and one that is not, then which lab would you choose?” And his lab and offices, which were originally going to be in the main biology building, were instead assigned by the university to a more remote, older building less vulnerable to attack. Meanwhile, Kreiter worries that the continued pressure and debates over the use of laboratory animals are harming German science. “It's not so much that established scientists are leaving Germany but that talented postdocs often stay abroad, in the U.S. or elsewhere,” he says. “Many of the postdocs just don't come back.” 13. SCIENCE AND SOCIETY # Getting a Measure of the Problem 1. Robert Koenig Statistics on animal experimentation in Europe as a whole are sketchy, but it is clear that the number of animals used for laboratory research in most countries has dropped substantially over the past 20 years. In Switzerland, the Federal Veterinary Office reports that the number of animals used for experiments declined from 2 million in 1983 to 510,000 in 1996. In Germany, the numbers dropped 40% between 1989 and 1997 to less than 1.5 million. And British researchers used 2.6 million laboratory animals in 1997—about half the number of lab animals used in the 1970s. Europe-wide, a new method for generating European Union statistics on laboratory animal use was agreed on in 1997, and the first set of new statistics should be published in a year or so. The downward trend is in part due to increasingly strict laws governing animal experimentation in northern and central Europe. Although regulations and procedures vary greatly from country to country and are not directly comparable, Mark Matfield, director of the European Biomedical Research Association, says Britain's law is the strictest, with personal licenses for research requiring time-consuming permission from both local and national authorities. “It can take up to a year” for British researchers to get such approval, Matfield says. Aside from the United Kingdom, researchers consider Switzerland and Germany to have Europe's strictest laws governing animal research. According to Ivar Aune, director of Germany's main lobbying group on behalf of researchers, “the Swiss laws may be stricter, but German officials seem to be more demanding in enforcing the laws.” Although rigorous, German regulations were altered last year and now require local boards to make the decisions on applications to use lab animals within 3 months. Despite the tightening rules surrounding the use of lab animals, many researchers who spoke with Science felt that the downward trend in animal numbers is leveling off and may begin to rise slightly again. Contends Andreas Kreiter of the University of Bremen: “All possible compromises have been made, of course always to the detriment of science.” 14. BIOPHYSICS # New Clues to Why Size Equals Destiny 1. Dana Mackenzie* 1. Dana Mackenzie is a science writer based in Santa Cruz, California. Dueling theories aim to explain why larger organisms tend to live longer; both conclude that the physics of nutrient distribution is key Turning 50 inevitably gets you thinking about how much time you have left in the world. For Geoffrey West, reaching the half-century mark in 1990 set him grappling with an extraordinary equation that predicts how long an organism might live: the quarter-power scaling law. In general, the larger the species, the longer the life. This relationship holds true with remarkable precision: Life-span tends to lengthen—and metabolism slows down—in proportion to the quarter power of an animal's body weight. “If you could understand the origin of these scaling laws,” says West, a high-energy physicist at Los Alamos National Laboratory in New Mexico, “you'd understand something about aging and death.” One of the few all-encompassing principles in biology, the quarter-power law helps scientists work out, for example, how to adjust drug doses in rats for use in humans. But the law's universality is baffling: Why should so many species, with their variety of body plans, follow the same rules on longevity? “Either there is some fundamental physical reason for it,” says Peter Dodds, a geophysicist at the Massachusetts Institute of Technology, “or evolution has found only one of many mechanisms that could work, and this has pervaded all of biology.” Physicists, of course, would prefer a physical explanation, and there are signs that they could be homing in on one. On page 1677, West and his colleagues propose that the quarter-power law derives from the physical constraints of an ideal system for distributing nutrients, whether it's the blood conduits running through our body or the vascular network nourishing a plant. An alternative theory from a second group (Nature, 13 May, p. 130) differs in details but comes to the same conclusion: The constraints on life-span and metabolism lie within an organism and have nothing to do with outward size and shape. Late last century, biologists sought to explain why smaller animals spend life in the fast lane and die young, while larger ones burn energy more slowly and live longer. Grab a chicken, and you will feel its pulse racing at about 300 beats a minute. Sidle up to an elephant about 10,000 times as massive as a chicken, and the thumping clocks in at 30 beats per minute. Because nearly all mammals expire after anywhere from 1 billion to 2 billion heartbeats, an elephant, naturally, should outlive a chicken. In 1883, biologist Max Rubner proposed an explanation. If an animal is N times as big (in height or length) as another, then its skin surface should be N2 times as big, he argued, and its mass (M) N3 times as big. Because the heat an animal can shed is proportional to skin surface, its total metabolic rate—the energy an animal burns in an hour—is proportional to N2, which is itself proportional to M2/3. Finally, the specific metabolic rate—the energy burned per unit mass, which controls pulse rate—is obtained by dividing by M, giving M−1/3. Thus, Rubner concluded, specific metabolic rate should decrease with size as the cube root of mass. This argument would work if chickens and elephants were spheres. Alas, they are not. Although the cube-root law does sometimes hold when comparing metabolisms of individuals in a species or among closely related species, it fails when extended to a divergent set of species. In 1932, Max Kleiber, a Swiss-American animal scientist, plotted the first accurate measurements of size versus metabolic rate, discovering that the correct scaling law was nearer to the fourth root. (The exponents he measured were 0.74 for total metabolic rate, and −0.26 for specific rate.) Thus it is no accident that an elephant, with 104 times the mass of a chicken, has a pulse rate about 1/10 as fast. Kleiber's law has been confirmed in a broad spectrum of animals (see chart). Why it's a quarter power and not the cube root is the mystery of life that West was pondering when he met biologists Jim Brown and Brian Enquist of the University of New Mexico, Albuquerque, who were turning the same question over in their minds. Enquist had found that the quarter-power scaling law holds for metabolism in plants, which deepened the intrigue. If there were a single explanation, they reasoned, it had to involve a feature shared by species in both kingdoms. Their suspicions centered on the circulatory system. In animals and plants, circulatory systems resemble branching fractal networks, and capillary size does not depend on organism size: An elephant's capillaries are the same size as a chicken's, and a pine tree's vascular capillaries are the same size as a daisy's. In 1997, the New Mexico trio published a report in Science (4 April 1997, p. 122) deriving the quarter-power scaling law from the physics of capillaries and the hydrodynamics in tubes. The model predicts accurately other scaling in mammals: aorta size, capillary density, and heart size, to name a few. “It's phenomenal how well it works,” West says. The revival of the dormant debate over Kleiber's law fascinated physicists. “It really was a paper with a brilliant idea,” says Dodds. One enthusiast was Jayanth Banavar, a physicist at Pennsylvania State University in University Park. He heeded Occam's razor and looked for a less complex rationale for Kleiber's law. “When you see something so pervasive, the explanation had better be really, really simple,” he says. Banavar found a way to take fractals, which he saw as an unnecessary complication, out of the picture. In last month's Nature paper, he and two Italian physicists—Amos Maritan and Andrea Rinaldo—present the case that the quarter-power law is a feature of any optimally efficient network. They assumed that a network for distributing nutrients has circulation length L, which in three dimensions serves roughly L3 sites where nutrients leave the system. Any nutrient would, on average, pass L sites on the way to its destination, they figured. Therefore, the total nutrients in the network at any given time must be roughly L4 (destinations times stops en route). This represents blood volume, which based on empirical observations should be proportional to an organism's mass. The rest of the argument is reminiscent of Rubner's. The total metabolic rate is proportional to sites served (L3, or M3/4), which makes the specific metabolic rate proportional to M−1/4, in agreement with Kleiber's law. The argument can also be adjusted to two-dimensional river networks, where it predicts cube-root scaling, in agreement with experimental observations. Some experts say the simplicity of Banavar's model is a big plus. “I'd be happier if it were this way than the fractal model,” says William Calder, a biologist at the University of Arizona, Tucson. But it must be interpreted with care, adds West: For example, a “site” cannot be an individual cell, because this would imply that the proliferation of cells could not keep pace with increasing organism size. Banavar was not the only researcher trying to simplify the New Mexico team's model. West and his colleagues also felt that there had to be a broader argument that could apply to plants, animals, and even one-celled organisms lacking a vascular system. With that in mind, they scrapped the physics of fluid flow. In their latest Science paper, the team makes a case for a quarter-power law based mostly on geometry, particularly the hierarchical nature of circulatory networks. “We traded in the dynamics for the statement that hierarchy gives rise to a power law,” West says. They argue that an organism's “internal area”—the total area of its capillary walls—fills up space so efficiently that it, in effect, adds a third dimension. The “internal volume” of all the vessels feeding the capillaries adds an extra dimension as well, scaling as the fourth power of internal length. The distinction between internal and external lengths, areas, and volumes is crucial, West says. “You really have to think in terms of two separate scales—the length of the superficial you and the real you, which is made up of networks.” The new argument by West's team poses its share of head-scratchers: If the internal length of an animal's circulatory system increases less rapidly (M1/4) than the external length (M1/3), what happens when it becomes too short to reach the skin? The answer, Dodds believes, may be a delicate balancing act: “I would argue that there are different scales in different parts of the animal”—that is, not one internal length scale but several. In the end, simplicity may have its limits. “The final theory will not be as simple as Banavar's but may be around the level of what West has done,” Dodds says. “What I am sure of is that they will both be very controversial.” 15. FLUID DYNAMICS # Soap Films Reveal Whirling Worlds of Turbulence 1. Robert Irion* 1. Robert Irion is a science writer in Santa Cruz, CA. New techniques for studying “soap film tunnels” are giving researchers a glimpse into the surprising world of two-dimensional turbulence Blowing a killer soap bubble takes a steady hand and a still day, as any child knows. But creating the kinds of soap films that thrill physicists goes far beyond child's play. Their high-tech films, clinging to a pair of wires hung a few centimeters apart, flow toward the floor and stream past strategically placed obstacles that stir roiling patterns in their wakes. The patterns, resembling swirls of smoke, are shedding new light on the intricate physics of turbulence in two dimensions (2D). “Soap films are a brilliant way to produce 2D turbulence,” says fluid dynamicist Patrick Tabeling of the Ecole Normale Superieure in Paris. “They are a major contribution.” Two-dimensional dances with whorls, scientists believe, arise in nature when moving liquids or gases are confined to flat or curved surfaces. Cyclones and other large-scale wind flows in the atmospheres of planets fit this bill, says physicist Walter Goldburg of the University of Pittsburgh. These vortices are hundreds of kilometers across but just a few kilometers thick, so their motions are essentially 2D, Goldburg says. Plasmas corralled by magnetic fields in space or in fusion reactors, and ocean currents pinned in a narrow horizontal layer by sharp changes in temperature or salinity, may also be examples of 2D turbulence. At first glance, some of the swirling 2D patterns in these systems look like cross sections of 3D vortices, the more familiar brand of turbulence generated, for example, behind a jet's wings. However, radically different principles of physics govern the two types of turbulence, says physicist Robert Ecke of Los Alamos National Laboratory in New Mexico. And although physicists have plenty of theoretical models and computer simulations of 2D turbulence, testing these models and theories in the lab has proven difficult. A turbulent flow initially confined to a 2D sheet likes nothing better than to unfold into three dimensions, as researchers trying to study 2D phenomena in wind and water tunnels discovered early on. But now, 2D turbulence is beginning to reveal its secrets, such as how its vortices evolve and suddenly swap energy, thanks largely to new techniques for studying “soap film tunnels.” The soap films build upon work in 1986 by French physicist Yves Couder, who dragged objects through stationary films to watch how turbulent patterns decayed to stillness. A few years later, a team led by fluid dynamicist Mory Gharib, then at the University of California, San Diego, pioneered a moving film that could be replenished continuously—a key advance that mimics flows in nature. Today, in a system devised by Goldburg and Xiao-lun Wu at Pittsburgh and refined by physicist Maarten Rutgers of The Ohio State University in Columbus, nozzles spray dilute dish soap continuously onto a pair of vertical wires. Cords draw the wires apart, stretching the soap into a rectangular patch of film that flows at speeds of a few meters per second. In the middle of the patch, the forces of gravity and air drag balance out so that the film reaches a steady “terminal velocity,” like a skydiver with outstretched limbs. The film, a sandwich of soap molecules and water, is several micrometers thick in the center and lasts for hours as pumps return the fluid to the nozzles at the top. Small cylinders inserted in the flow create vortices that look like mushrooms, spiral galaxies, and other graceful forms. These shapes are highlighted by microscopic spheres of titanium dioxide or polystyrene in the film, which catch the light of probe lasers or stroboscopic lamps. This system let Rutgers probe one of the key differences between 2D and 3D turbulence. In a 3D fluid, swirls can intensify in a process called “vortex stretching,” in which particles go faster and faster as they plunge toward the throat of the vortex. This happens to water gurgling down a drain or winds sucked into a tornado: The conservation of angular momentum amplifies motions as the whirling tube gets stretched thinner. However, this stretching and amplification can't happen in a 2D fluid, because the swirling particles are confined to a plane. Instead, according to theories proposed in the 1960s by physicists George Batchelor and Robert Kraichnan, some vortices should merge into progressively larger ones, a process called the inverse energy cascade. At the same time, smaller swirls should stretch and fold into the seething fluid like candy on a taffy puller—an effect known as the forward enstrophy cascade. To test these predictions, Rutgers placed two vertical combs of cylinders along the edges of his soap film flow to generate constant trains of interacting vortices as the film streamed past. He saw evidence for both processes. If the vortices exceeded a certain size, they merged into larger whorls. However, smaller vortices sheared and vanished, like thin ribbons of cream in a mug of coffee. “This was a key theoretical prediction, but [the two energy cascades] had never been observed simultaneously,” says Rutgers, whose findings appeared last September in Physical Review Letters. Ongoing work by Rutgers and by Goldburg and Wu suggests that 2D turbulence doesn't always conform to theory, however. Theory predicts that 2D vortices should grow and dissipate gradually, because they cannot interact as chaotically as they do in 3D. Instead, says Goldburg, “we are finding that 2D flows can depart violently from the mean in rare bursts”—a hallmark of turbulence called “intermittency” that scientists had assumed was the sole province of the 3D world. The 3D phenomenon is familiar to airline passengers as “violent excursions”—unexpected jolts that leave your stomach in the lurch. In 2D flows, the excursions appear as occasional spiky transfers of energy among vortices. Some experts wonder whether the soap film results are too squeaky clean. “The experiments are very nicely done, but they are not an exact analogy [to 2D turbulence] by a long shot,” says fluid mechanician John Lumley of Cornell University in Ithaca, New York. The main problem, he says, is that the films vary in thickness by up to 30%. Such bulges could be compressed or squished, especially in a fast-moving film, he says. To Lumley, that invalidates the assumption that the turbulence traces a purely planar flow, because compression introduces a 3D component of motion to the film. Experimentalists acknowledge that the bulges are a concern. But in a report last August in Physical Review Letters, Ecke and two colleagues described a study of a tilted film that flows more slowly than the vertical ones. They found that the behavior of their film, bumps and all, hewed closely to predictions for an ideal 2D sheet. Even so, Lumley wonders about the real-world relevance of soap films. “There is one important case where turbulence in nature is approximately 2D, and that is in the atmosphere on large scales,” he says. “Beyond that, my feeling is that almost all turbulence in the universe is three-dimensional.” Such views, however, shouldn't burst the bubbles of soap film researchers, says fluid dynamicist Katepalli Sreenivasan of Yale University in New Haven, Connecticut. “Finally, after 30 years, we see a convergence of theory, simulation, and experiment,” he says. “Anything that cleans up our understanding of one area of turbulence is most valuable.” 16. SCIENCE EDUCATION # Reinventing the Science Master's Degree 1. Mari N. Jensen* 1. Mari N. Jensen is a science writer based in Tucson, AZ. The Sloan Foundation is backing an experiment at five universities to offer science undergrads an alternative to the Ph.D. Meet Jarrell Pair, science professional. Fresh out of grad school, Pair is spending the next few months in Santiago de Compostela, Spain, creating a virtual driving tour of the 1200-year-old city's narrow cobblestone byways and bustling cafe-lined plazas. When the project is finished, people will be able to sit in a fake car surrounded by movie screens and take a virtual spin through Santiago, “experiencing all the sights and sounds of being in the actual city,” Pair says. The exhibit, part of the Santiago 2000 celebration, will also tour computer graphics conferences in Europe, serving as a showcase, Pair says, for luring high-tech companies to the Santiago region. Pair didn't have to tough out a Ph.D. for graduate training in his specialty, human-computer interaction. Instead, the 26-year-old with eclectic interests—he earned bachelor's degrees in computer engineering and international affairs, and certificates in music, drama, and film—last year completed a professional master of science degree at the Georgia Institute of Technology in Atlanta. The university is one of five in an experimental program, funded by the Alfred P. Sloan Foundation in New York, to test the academic waters by offering a graduate science degree that is tailored toward careers in business and industry rather than academe—an alternative to the long journey to a Ph.D. “I see this as an altogether new degree comparable to the creation of the MBA in 1908,” says Sheila Tobias, a science education consultant in Tucson, Arizona, who is advising Sloan. “No one will be able to manage anything without a firm grounding in science and mathematics,” says Tobias, who co- authored the book Rethinking Science as a Career. She envisions a cadre of technically trained, highly paid “science-trained professionals.” People in industry are gung-ho for the program. “There is a major demand for people with some practical experience other than the pure science,” says Jack Samiarias, senior vice president for ABN AMRO, an international bank in Chicago. But the new breed of science salaryperson must overcome several hurdles, including the disdain U.S. academic scientists generally shower on master's degrees. “Faculty in the sciences are mostly accustomed to thinking their students will get doctorates, and their students will get jobs in academia,” says Sloan program director Jesse Ausubel. Although a master's has long been a respectable pursuit in engineering, computer science, or geology, for example, it doesn't carry much weight in the basic sciences. Master's degrees in the life sciences, for instance, have evolved into “a consolation prize or booby prize for people who did not make it through qualifying exams,” says Shirley Tilghman, a Princeton University molecular biologist who chaired a panel on graduate education for the National Research Council (NRC). The situation is a bit different in other countries, particularly in Europe, where science students often earn a 6-year degree similar to a combined U.S. bachelor's and master's. In many fields, the master's used to be a required milestone on the road to a Ph.D. says John Vaughn, executive vice president for the Association of American Universities in Washington, D.C. Realizing that stopping at a master's “wasn't a real good deal,” he says, many departments eliminated it to streamline doctoral programs. Any university that starts a professional master's, Tilghman says, must “make this a distinct and different and new program.” Indeed, warns Vaughn, “programs like this will encounter resistance on some campuses that see it as resurrecting some second-class degree to pursue a third-class career objective.” But Tilghman and others say the time is ripe to renovate the science master's, particularly as many new Ph.D.s flounder in today's job market. By now, most scientists have disabused themselves of the notion that academia offers ample opportunities to freshly minted Ph.D.s. And some Ph.D.s who embark on careers outside academia find, to their dismay, that they didn't even need the years of grinding doctoral studies and late nights in the lab to succeed. As people began to grapple with this trend in the early 1990s, the idea of revitalizing the master of science started making the rounds. Georgia Tech got into the game early, conceiving the human-computer interaction program in 1995. Officials saw the new master's as “an interdisciplinary, industry-oriented, stand-by-itself degree—an alternative to the academic Ph.D.” says Anderson Smith, Georgia Tech's associate dean for the College of Sciences. Sloan officials also were exploring alternatives. “This was an idea that was in the air, and the time seemed right to do it,” says Ausubel. In early 1997 Sloan invited Georgia Tech and other universities to submit proposals for professional master of science programs. Georgia Tech won its Sloan award before launching its human-computer interaction program in fall 1997. Sloan has handed out four more$400,000 grants to universities to start degree programs that combine graduate science education with training in business-related fields such as finance, computation, and information sciences. So far 19 Sloan-funded programs, spanning fields from computational linguistics to integrated pest management, are up and running or in the works (see table). Other universities, not funded by Sloan, have started or are contemplating similar programs. All are designed to appeal to students who don't want to become steeped in the esoterica of subdisciplines.

The Sloan-funded programs last up to 2 years and focus on coursework rather than basic research. Like MBA programs, the science master's emphasizes hands-on experience through internships in industry. Tuition and fees per semester range from $1132 to$11,099; the fledgling programs are providing some financial support to lure students. But in the long run, officials say, students in these professional M.S. programs will be expected to foot the bill themselves, unlike Ph.D. students, who generally receive training grants or stipends to cover tuition.

View this table:

Industry experts say they are thrilled with what the new programs are aiming to do. Ph.D.s, they say, often are too specialized and have trouble adapting to nonacademic work. Robert Arnott, managing partner for First Quadrant, an investment management firm in Pasadena, California, hires Ph.D. physicists as financial analysts. However, he says, most recruits “start out not knowing what to do. We wind up spending the first year, or sometimes two, helping them get out of blind alleys.” He's curious how the science master's students will fare and is eager to bring some aboard as interns.

Educators, too, are keen to see the effort blossom. “There is a niche for professional master's degrees,” says Wyn Jennings of the National Science Foundation's Division of Graduate Education. But first, he says, academics must discard their disciplinary blinders. “You will have to divorce yourself from traditional disciplines,” he says, because the professional master's programs are driven by the needs of the market, not specific fields.

If the programs earn respectability and take root, they “would save a lot of young people a lot of angst,” says the NRC's Charlotte Kuh. Pair certainly seems angst-free: He says he has received many “unsolicited inquiries as to my availability” after his return to the United States in August.