# News this Week

Science  27 Oct 2000:
Vol. 290, Issue 5492, pp. 682
1. MARINE ECOLOGY

# Warmer Waters More Deadly to Coral Reefs Than Pollution

1. Dennis Normile

NUSA DUA, INDONESIA—Contrary to conventional wisdom, warmer ocean waters are a greater threat to coral reefs than local environmental insults. That assessment comes from a new scientific report* released this week that documents a sudden and steep jump in damage stemming from the 1997–98 El Niño-La Niña event.

“Coming up to 1998, we thought direct human impacts were the biggest threat to reefs,” says Clive Wilkinson, editor of the report and a marine biologist at the Australian Institute of Marine Science in Cape Ferguson, referring to such events as pollution and destructive fishing practices. “Now, we've got to reset our agenda to focus on climate change as well.” He admits that climate change was first mentioned as a possible cause in the 1980s, but the incidence of bleaching then was isolated and at a low level. “In the latest event, we had bleaching all the way from Brazil to the Indian Ocean,” Wilkinson says.

The new analysis comes from the Global Coral Reef Monitoring Network, an affiliation of marine scientists and government agencies that 2 years ago produced its first snapshot of the health of reefs. Presented here at the 9th International Coral Reef Symposium in Bali, the current report is the first that contains a complete description of the physical damage caused by warmer than usual sea temperatures during 1997 and 1998. Prior to 1998, an estimated 11% of the world's known reefs had been destroyed by human activities, according to the report. Barely 1 year later, another 16% had been “severely damaged” by the El Niño event, with little chance of short-term recovery (see table). Although some of these reefs will recover over time, others have already slipped into the destroyed category. “It was a wake-up call for reef scientists,” says Wilkinson, noting that climate change models predict not only a steady rise in baseline sea temperatures but possibly more frequent and ferocious El Niños that would cause more bleaching.

View this table:

Often called the “rainforests of the sea,” coral reefs host a disproportionate share of marine life. What's more, “their sensitivity to rising temperatures” also makes them “silent sentinels” of global warming, says Alan Strong, an oceanographer who studies global climate change at the U.S. National Oceanic and Atmospheric Administration in Camp Springs, Maryland. “We don't have many systems that give us such a characteristic climate change signal as coral reef bleaching,” Strong says.

Even small temperature changes can have dramatic effects on reefs. Exceeding a threshold of around 30°C triggers bleaching, in which coral expel zooxanthellae, the symbiotic algae that give the coral its color. Coral appear able to survive short-term bleaching if there are no other stresses, such as pollution or destructive fishing. They can die, however, if subjected to high temperatures for long periods of time or if already stressed. Previous bouts of bleaching have been limited to particular areas, such as the Caribbean. But the 1997–98 El Niño event, which lasted for over a year, produced record-high sea surface temperatures throughout the Indian and western Pacific oceans; this in turn led to the most extensive coral bleaching ever seen. “Some of those reefs will come back, but others won't,” Wilkinson says.

The intensive studies of this bleaching event, many being reported for the first time at the symposium, are likely to shed light on which reefs are at risk from bleaching, the mechanism of bleaching, and the possibilities of recovery. Reefs in areas with weak ocean currents or those already damaged by human impact fared the worst. But even populations thought to be in ideal health, like those around the Maldive Islands in the Indian Ocean, succumbed to the bleaching. Alternately, reefs washed by strong currents, which typically bring cooler water up from the depths, were often spared.

The news is not entirely bleak. Some reefs have recovered from the 1997–98 event more quickly than researchers expected, says Terry Done, a senior research scientist at the Australian Institute of Marine Science and the current president of the International Society for Reef Studies. In some cases, coral apparently survived deep within reef structures; in others, new coral was recruited from nearby deeper, cooler water. Several groups are also reporting that certain species of coral can apparently adapt to higher temperatures. But Done warns that these findings should not lead to complacency. “It's a question of the quality of the reefs in the short term,” he says. Even those reefs showing early recovery face a decades-long process to regain their previous state. “And while there is some adaptive capability, it is unlikely to be at rates fast enough to cope with global warming,” he says.

One bright spot, say Wilkinson and others, is the growing interest of tropical nations, where the majority of the world's reefs are located. Indonesia has started to address the local human impact on the reefs through a coral reef recovery program that combines education with more careful management. But “climate change is a tragedy not of our making,” says Sarwono Kusmaatmadsa, Indonesia's Minister of Maritime Affairs and Fisheries, calling on the industrialized world to shoulder more responsibility for the causes of global warming. Although potentially costly to industrialized nations, such efforts would be good for local economies, conferees noted. Indeed, coral reefs provide the basis for an estimated $400 billion fishing and tourism industry around the world. That figure provides another—and for some more compelling—reason to protect coral ecosystems. • *“Status of Coral Reefs of the World: 2000.” 2. 2001 BUDGET # NSF and NASA Score Last-Minute Victories 1. Andrew Lawler, 2. Jeffrey Mervis Moved to generosity by the impending elections and a big budget surplus, Congress last week gave both NASA and the National Science Foundation (NSF) significant hikes for 2001. After traveling a rocky road to reach this point, legislators gave NSF$4.42 billion, a $522 million boost over this year that nearly matched NSF's 17% request. NASA received$14.3 billion, nearly twice the White House's request for a 3% boost—but with hundreds of millions of dollars in earmarks added on.

When the House and Senate differ on funding, they usually produce a final budget by splitting the difference. But this year leaders “compromised” on a total for both NSF and NASA that exceeded the earlier levels set by either body. “I really like Congress's math this year,” quipped NSF director Rita Colwell. “I'm thrilled with the outcome.” Leaders greased the legislative process by adding in numerous last-minute increases requested by such key members of the Appropriations Committee as Senators Barbara Mikulski (D-MD), Robert Byrd (D-WV), and Ted Stevens (R-AK).

The Senate had been considering a NASA bill nearly $200 million below the Administration's request, which would have required the space agency to scale back many programs (Science, 22 September, p. 2018). The House version was lower, at a whopping$377 million less than the request and just slightly above the 2000 level. The final bill, however, leaves space science with a $2.5 billion budget—$100 million more than requested and well above the $2.2 billion spent in 2000. Ed Weiler, NASA's space science chief, cautions that the boost won't give him much wiggle room to cope with inflation in planetary missions, several of which are likely to cost more than promised. The flexibility disappeared because much of the new money will go to pork-barrel projects, such as$10.5 million for education centers on Mauna Kea in Hawaii, $4 million for a visitor center at the Green Bank Radio Astronomy Observatory in West Virginia, and$2 million for equipment at the South Carolina State Museum's observatory, planetarium, and theater in Columbia. But Weiler is trying to borrow funds from a planned mission to Jupiter's moon Europa to keep one project—a flight to Pluto—from a lengthy delay. Weiler, who aims to rule on the Pluto mission by the end of November, acknowledges its scientific merit but notes that “Europa is clearly the priority of the White House.”

In contrast to the small increase NASA requested, NSF asked for a record $675 million boost in 2001, or 17%. In June the House voted for a rise of just 4%, and last month the Senate approved a 10% hike, so the final 13.3% boost made NSF officials very happy. Even so, Congress failed to fully support several key initiatives. The bill provides$215 million of a $327 million request for information technology research,$150 million of the $217 million sought for nanotechnology, and$75 million of the $136 million planned for biocomplexity. But Congress responded with enthusiasm to projects that promised tangible benefits for local institutions and had strong backing from influential sectors of the scientific community. Although legislators avoided earmarks to individual institutions, they shelled out more than the Administration had requested for programs that support smaller states, graduate fellowships, and informal science education. They also rejected NSF's request for$29 million to begin two ground-based research networks, substituting $12.5 million to continue work on a high-altitude research plane that had fallen off NSF's list of priorities for 2001. And they added$15 million for badly needed upgrades and repairs to radio telescopes in West Virginia, New Mexico, and Puerto Rico.

One big winner is the agency's 20-year-old program to bolster the 20 states that traditionally receive the fewest federal research dollars. Long a congressional favorite, the Experimental Program to Stimulate Competitive Research (EPSCoR) this year received a 56% boost, to $75 million. “Everybody's delighted,” says Joe Danek, head of the nonprofit EPSCoR Foundation that represents the eligible states. The money will help NSF fund a competition now under way that will award up to$3 million a year to build research capacity in EPSCoR states and assist researchers applying for funding through regular channels.

President Bill Clinton is expected shortly to sign the bill, which was bundled with a $24 billion measure to fund the Department of Energy and various water and conservation projects. 3. ACADEMIC COMMUNITY # Institute Goes to Court to Remove Researcher 1. Constance Holden The Institute for Advanced Study (IAS) in Princeton, New Jersey, has served for 70 years as a peaceful haven for scholars, including Albert Einstein. But this fall it is embroiled in an uncharacteristically tense—and public—fight to remove one of its tenured professors. The persona non grata is Piet Hut, a 47-year-old astrophysicist who was hired at the precocious age of 32. This summer the institute asked a court to enforce a 1996 agreement in which Hut promised to leave by 2001. This month Hut countersued, saying that he was coerced into signing the document and that the institute is trampling on his academic freedom. The matter was first reported by The New York Times. A number of scientists have come to Hut's defense; many feel that the institute is making a big mistake. “It's not Piet Hut's fault,” says University of California, Berkeley, astronomer Frank Shu, who adds that “the institute made a gamble” when it appointed someone so young. “Now they have to either live with it or find some compromise.” Alar Toomre, an applied mathematician at the Massachusetts Institute of Technology, says that IAS, by going to court, “is damaging its own reputation more than anything Piet Hut might have done.” The institute's director, Phillip Griffiths, issued a statement saying that the conflict is a contractual one and “not an issue regarding tenure or academic freedom.” In its suit, filed on 25 July before the federal district court of New Jersey, IAS argues that Hut hasn't fulfilled his early promise and should abide by the 1996 agreement. Institute officials declined to comment further on the case. IAS tapped Hut in 1985 as an up-and-coming assistant professor at Berkeley with a solid publication record in stellar dynamics. In 1986 he and an IAS postdoc, Joshua Barnes, published the famous Barnes-Hut “tree algorithm” that is widely used in computer simulations. By 1989, however, according to the institute's complaint, then-director Marvin Goldberger felt that Hut ought to “look for another position,” and Hut agreed that “he was not performing … and … would never perform at the level … achieved by other faculty members.” In 1993, a visiting committee called Hut's appointment a “mistake,” according to the court document, and in 1995 the faculty of the School of Natural Sciences agreed at a meeting—attended by Hut—that “Hut's presence was having a detrimental effect” on the IAS. In 1996, the institute froze Hut's salary and got him to sign a letter agreeing to leave in 2001. In 1999, he signed a formal contract, but withdrew his consent during the 1-week period allowed by the contract. Hut dismisses the negative job assessments, saying that there has been no formal evaluation of his work and that the visiting committee had little regard for his field of computational physics. He says the problem started with a 1993 dispute with string theorist Ed Witten over Hut's desire to buy an expensive supercomputer. Witten has declined to comment. Hut also says that the institute threatened to cut his salary and marginalize him if he didn't sign the 1996 letter. He says he went along initially because he saw no alternative. Hut has no shortage of supporters. Computer scientist Joseph Traub of Columbia University and the Santa Fe Institute calls him “one of the most stimulating, creative, intelligent, serious scientists” he has known. More than a score of other scientists, including Princeton astrophysicists Edwin Turner and Bohdan Paczy'nski, have publicly defended Hut's scientific credentials and accused the IAS of violating his academic freedom. Hut also has many admirers among those working in the interface of science and religion, a subject that has attracted his interest in recent years. As for his earlier work, Shu explains that stellar dynamics and computational modeling was a hot area in 1985 but that, “for whatever reasons, the subject suffered a decline” as cosmology and theoretical physics moved to the fore. Some of Hut's contributions, such as building a special-purpose chip for rapid calculations, are more valued in engineering than in physics, he adds. Whether or not Hut belongs at IAS, many scientists think that the institute has committed a major blunder in going to court. “They're giving themselves and, to some extent, science a black eye,” says Shu. “It's bad for everybody.” 4. ECOLOGY # Pacific Salmon Run Hot and Cold 1. Kathryn Brown* 1. Kathryn Brown is a writer in Alexandria, Virginia. Overfishing. Dams. Disease. There's plenty to blame for the ups and downs of Pacific salmon—and humans are often the guilty party. But a new study spotlights another, more natural force driving salmon numbers: climate. Using a novel technique, described on page 795, paleoceanographer Bruce Finney of the University of Alaska, Fairbanks, and his colleagues have been able to chart the abundance of sockeye salmon in the Bristol Bay and Kodiak Island regions of Alaska over the past 300 years—by far the most complete record yet. Through time, they found, sockeye populations have alternately soared and slipped, following natural climate variations—well before commercial fishers began throwing nets over the sides of boats. “This paper is terribly important, because it's the first solid proof that salmon abundance fluctuates naturally,” says ecologist Richard Beamish of Canada's Department of Fisheries and Oceans. Although many researchers have assumed that climate variability affects Pacific salmon, precise data have been scarce. “Now, we have sound scientific evidence that climate is a natural part of salmon dynamics,” Beamish says. And that could change the way salmon stock managers set harvest limits or evaluate fish population swings, he adds—particularly in the beleaguered Pacific Northwest, where many sockeye stocks have already gone extinct or live at record lows (Science, 4 August, p. 716). The new study also highlights an early factor vital to the survival of salmon before they set out for sea: the nutrient-rich carcasses of adult fish that have migrated back home to nursery lakes. Finney and colleagues studied sockeye salmon near pristine Kodiak Island and Bristol Bay on Alaska's southern coast. There, millions of sockeyes hatch and live in lakes for up to 3 years before migrating, during spring, to the Pacific. After swimming a well-worn sea route for 2 or 3 years, the salmon return home, spawn, and die. While at sea, sockeyes feast on plankton, squid, and small fish, packing on 99% of their adult body weight. And it's this sea buffet that makes the salmon easy to track decades later. Consuming vast amounts of seafood, the fish take on a high ratio of nitrogen-15, a stable nitrogen isotope, before returning home. Few other fish or atmospheric factors add large amounts of nitrogen-15 to freshwater lakes, so the isotope—which eventually settles, year after year, in the layers of the soupy lake floor—is a tracer for past salmon populations. “We can recreate history from records left behind in the sediment,” Finney says. Finney's team decided to reconstruct history in five Alaskan salmon nursery lakes: Karluk, Red, and Akalura lakes on Kodiak Island, and Becharof and Ugashik lakes in nearby Bristol Bay. They included two additional lakes as controls. From float planes or boats, the researchers sunk an 18-kilogram core—basically, a barrel with weights—into the bottom of each lake and retrieved a sediment sample. Every half-centimeter of the green-brown goop corresponded to about 5 years of lake history, according to time markers such as volcanic ash in the sediment. The scientists measured nitrogen-15 in each section of lake sediment, estimating changes in the number of salmon that returned to the lake over 3 centuries. Then, they compared those salmon population trends with fishing data from commercial catch records and with past climate descriptions, based on previously published tree-ring data. The emerging pattern was clear: Alaskan sockeye numbers tended to drop when ocean temperatures fell and jump when waters warmed. In the early 1800s, for instance, the sea surface temperature was about 1°C below average, and an estimated 500,000 salmon returned to Karluk Lake every year. By the 1920s, however, the sea surface had warmed by 1.5 degrees—and the number of returning salmon had more than quadrupled to an estimated 2.5 million fish. “We can see this distinct variability happening on the order of 50 to 100 years, with climate change driving salmon populations up and down in a coherent way over fairly large regions,” Finney says. By contrast, control lakes, without salmon, showed consistently small amounts of nitrogen-15. The sediments also recorded the collapse of Alaskan salmon with the onset of fishing in the 1880s. And it wasn't just the number of netted fish that led to the sockeye's decline. It was also, Finney says, the drop in nursery lake nutrients when fewer mature salmon returned to spawn and die. Normally, salmon carcasses litter the lake floor, releasing nitrogen and other essential nutrients consumed by plankton and, in turn, young salmon. Along with nitrogen-15, Finney's team counted plankton fossils trapped in lake sediments. In years when many salmon returned to the lake, eutrophic algae and zooplankton thrived. Today, Alaskan salmon enjoy a favorable climate. More careful management—and a warming trend—boosted the sockeye to record highs in the 1990s. But farther south, those warmer waters spell trouble. While salmon in the Gulf of Alaska have prospered in nutrient-rich, warm waves, salmon numbers in the Pacific Northwest have dropped. There, researchers suspect, warming waters tend to stratify differently than in the north, trapping nutrients too far below the plankton and migrating salmon in the upper ocean. The warmer coastal ocean has packed an added wallop for Pacific Northwest salmon, already stressed by overfishing and habitat loss, says Nathan Mantua, a climate scientist at the University of Washington, Seattle. The new study, Mantua says, underscores the challenges facing the flashy red sockeye both at sea, in a fickle climate, and at home, where vital lake nutrients vary. “The lesson is that we really can't ignore what happens at any stage of the salmon's life cycle,” Mantua says. Salmon management, he adds, may grow even more important in the future. Waters in the Pacific Northwest appear to be cooling, but the specter of global warming looms. As Finney puts it: “We've still got a lot to learn.” 5. RESEARCH COSTS # Canada to Begin Funding Overhead on Projects 1. Wayne Kondro* 1. Wayne Kondro is a writer in Ottawa. OTTAWA, CANADA—Winning a competition last year for Canada's biggest new scientific facility in 30 years was quite a coup for the University of Saskatchewan. But along with the right to host the$116 million Canadian Light Source (CLS) came a king-sized headache: the need to find an estimated $9.4 million each year to operate the 2.9-giga electron volt third-generation synchrotron radiation facility once it's finished in 2004. It's a problem facing all Canadian universities, which unlike their U.S. counterparts receive no money for overhead on federally funded research projects. But help may be on the way. Last week federal Finance Minister Paul Martin announced a$268 million outlay for future equipment awards provided by the Canada Foundation for Innovation (CFI), a $1.3 billion entity created in 1997 to rejuvenate labs in universities and research hospitals. The funds would be awarded competitively in support of infrastructure grants such as the$37.8 million that CFI gave Saskatchewan last year to help finance the synchrotron facility.

The money, part of an unusual minibudget unveiled in the run-up to a parliamentary election scheduled for 27 November, is the first direct federal outlay for overhead costs, which up to now have been met by a combination of provincial operating grants to universities and federal transfer payments for postsecondary education. School administrators say it meets a desperate need.

“Universities have to be able to take advantage of research opportunities,” says Peter MacKinnon, president of the University of Saskatchewan. But big projects like the CLS do carry unreimbursed costs, as do individual investigator grants awarded by the country's three research granting councils. “Universities are often left with the obligation of finding money for matching programs or for meeting indirect costs,” MacKinnon says. “It's imposing a very considerable burden on all institutions, particularly on the research-intensive ones.”

To fill the gap, administrators have traditionally looked to private donations and grants from provincial governments. But those sources are drying up, says Manuel Buchwald, chief of research at the Hospital for Sick Children in Toronto. As provincial governments clamp down on health care spending and on university budgets, he adds, “it's becoming increasingly difficult for the institution where the research is done to provide the indirect costs.”

The new pot of money is a response to those pressures. But it won't erase the problem. The money can't be used to support existing equipment or facilities, notes CFI president David Strangway, and it won't be given automatically to all future CFI infrastructure awardees. “There are a lot of interesting questions for us to resolve,” says Strangway. For example, he says, CFI has yet to come up with a good definition of indirect costs, and it's still debating whether small replacement facilities should get overhead funding.

Although the logistics must still be worked out, Strangway imagines asking groups seeking future CFI infrastructure grants to include a specific request for indirect costs. These combined proposals would be assessed by peer-review committees. This new process will start in the next funding cycle, which begins in January with a call for proposals, leading to awards in late 2001.

MacKinnon sees a plan for growth in that next cycle. The light source, called Sas-katchewan's “Field of Beams,” is scheduled to open with 10 beamlines (conduits for carrying the synchrotron light to workstations), and researchers hope to add 20 more. MacKinnon says that “it would be nice to think that CFI operations funding would be available for additional beamlines.”

University officials say they welcome the support but note that the new fund only scratches the surface of what is needed. They hope for additional resources after the election from the governing Liberals, who have a commanding 20% to 25% lead in the polls. “There is still a very, very compelling case for a broader program to deal with the indirect costs of tricouncil-funded research and other publicly funded research,” says MacKinnon.

6. PLANETARY SCIENCE

# Saturn Wins Satellite Title With New Moons

1. Richard A. Kerr

This week an international team of astronomers announced the discovery of four new moons of Saturn, restoring the ringed planet to its status as commander of the largest retinue of satellites in the solar system. Their appearance should help researchers understand not just how the new moons were formed but also how the giant planets themselves came to be.

At this week's meeting of the Division for Planetary Sciences in Pasadena, California, astronomer Brett Gladman of the Observatory of Nice and seven colleagues reported that state-of-the-art light detectors (see following story) revealed four bodies 10 to 50 kilometers in diameter that “almost certainly” are orbiting Saturn. That makes a total of 22, surpassing Uranus's 21.

Although their orbits have not been determined yet, the new moons are probably “irregular” satellites. Whereas most major satellites form from dust and gas orbiting a planet, irregulars are outsiders captured by a planet into distant, inclined, and sometimes highly elongated orbits.

Two groups of four irregular moons orbit Jupiter in opposite directions. Astronomers take that arrangement as a sign that two large bodies approached a still-growing Jupiter, broke into pieces, and went into orbit. With four more examples to study, astronomers may be able to choose between two theories of how precursors of irregular satellites were captured. One model involves close encounters or even collisions with existing moons; the second points to drag from the last wisps of gas accreting to a new planet.

7. PLANETARY SCIENCE

# Solar System Scientists Look to Find an Edge

1. Richard A. Kerr

The universe may go on forever, with no end or edge in sight, but our solar system is hardly so expansive. For several years, ever-improving telescope technology has allowed astronomers to peer farther and farther beyond Neptune to discover a rapidly increasing number of bodies littering the outer reaches of the solar system. Now many researchers agree that an end is in sight, although some remain skeptical.

A report this week at the annual Division for Planetary Sciences (DPS) meeting in Pasadena, California, places the limit just beyond little Pluto's farthest wanderings, about 50 times Earth's distance from the sun (50 astronomical units or AU). Although some objects that originally formed in the solar system at its birth 4.5 billion years ago have been flung beyond that point, astronomers can see none beyond the 50-AU limit still hanging around their birthplace. An edge there would be far closer than predicted by conventional theories of solar system formation, suggesting that our planetary system started out surprisingly small or something ripped away the outer parts of the nascent solar system.

The solar system “really stops beyond Neptune” and the chunk of primordial debris called Pluto, says astronomer Brian Marsden of the Harvard-Smithsonian Center for Astrophysics in Cambridge, Massachusetts. “A lot of us have concluded that”—a lot, but not everyone. “This is a tricky game,” says astronomer Brett Gladman of the Observatory of Nice. “When I look at our data, I don't think there's any real evidence for an edge” beyond which there is little or no detectable material. Although researchers gather their evidence for an edge from their own deep telescopic surveys, proof may have to await a close comparison of existing surveys, possibly in the next year.

Putting a close-in edge on the solar system would have more than merely territorial implications. The extent of the Kuiper belt, a disk of bodies left over from the primordial gas, ice, and dust that agglomerated to form the planets, can provide hints of how the planets came together. An edge at 50 or 55 AU would mean the original preplanetary disk was far smaller than those of young stars seen today, or that something—perhaps a too-close encounter with another newly formed star—tore away much of the disk.

Proponents of an edge place it at the apparent outer limit of today's Kuiper belt. The young giant planets Jupiter, Saturn, Uranus, and Neptune gravitationally tossed great balls of primordial stuff far outward, where they formed the Oort cloud tens of thousands of AU from the sun. Nevertheless, 4.5 billion years later, upward of 100,000 objects at least 100 kilometers in diameter still orbit in the Kuiper belt beyond Neptune, according to extrapolations from telescopic surveys.

Just how far Kuiper belt objects (KBOs) extend beyond Neptune has long been a mystery. But in the past decade astronomers have used larger versions of their workhorse light detector—the charge-coupled device—and assembled them into large mosaics. With these far more capable CCDs, planetary astronomers are detecting fainter and fainter objects while surveying larger parts of the sky.

Until recently, over 300 KBOs had been found, but none much more distant than about 50 AU. At the DPS meeting, astronomers Lynne Allen and Gary Bernstein of the University of Michigan, Ann Arbor, and planetary dynamicist Renu Malhotra of the University of Arizona in Tucson reported that their state-of-the-art survey has discovered 24 new KBOs, but none was found beyond 53 AU even though their search was sensitive enough to find 160-kilometer objects out to 65 AU. And astronomers Chadwick Trujillo of the California Institute of Technology in Pasadena, David Jewitt of the University of Hawaii, Manoa, and Jane Luu of Leiden University in the Netherlands reported that their search over a much larger area for less-faint objects than Allen and colleagues could detect turned up 57 KBOs, none of which is beyond 50 AU. “It seems like there's an edge,” says Trujillo. Allen agrees: “We see objects nearby, and we don't see anything far away.”

That finding holds for other surveys, too, although Gladman cautions that looks could be deceiving. “I don't think the evidence is strong,” he says about the new results. “It's very hard, because these objects become faint very fast with increasing distance. And the results depend on the assumptions used to model the data.” Those assumptions may be spelled out in the next year, says Gladman, leaving astronomers in a better position to judge whether the solar system ends abruptly.

8. EASTERN EUROPE

# Help Needed to Rebuild Science in Yugoslavia

1. Richard Stone*
1. *With reporting by Robert Koenig in Croatia and Slovenia and Gillian Sandford in Belgrade.

The promise of democratic rule in Yugoslavia has rekindled scientists' hopes that their careers will eventually get back on track and that scientific cooperation between the former Yugoslav republics can be reestablished

BELGRADE—A sprawling nuclear lab on the Danube 15 kilometers downstream from this emotionally spent capital offers a window on this beleaguered nation's scientific past—and on the hopes of Serbian scientists that they may now have a future.

During the Cold War years, the Vinča Institute of Nuclear Sciences was a bastion of top-notch research, and as Yugoslavia warmed to the West in the late 1970s, foreign physicists beat a path to its door. Funding declined during the 1990s, but even as it was drawing worldwide condemnation for its role in the war in neighboring Bosnia-Herzegovina, the Serbian government in September 1992 broke ground on a state-of-the-art cyclotron facility at Vinča for everything from probing atomic structure to treating cancer patients.

By 1998, the government had cobbled together enough funding to allow three key components—a machine that churns out hydrogen ions, another for heavier elements, and a channel for blitzing targets for materials science experiments—to come on line. “It's the best ion source [of its class] operating in the world today,” beams Dejan Trbojevic, a former Vinča physicist now at Brookhaven National Laboratory in Upton, New York. All that remained unfinished was work worth about $5 million on the centerpiece, the VINCY Cyclotron, and two more experimental channels. The entire TESLA Scientific Center, named after Yugoslavia's homegrown star from a century ago, electrical engineering wizard Nikola Tesla, was nearly complete. But suddenly it came to a screeching halt. In 1998, Serbia's newly appointed science minister, Branislav Ivković, without explanation pulled the plug on TESLA's financing—a move that Vinča scientists, who draw salaries of about$50 per month from the institute's tiny budget, were powerless to offset. United Nations and European Union sanctions, in place since 1992, ruled out a Western bailout. Hopes for the facility almost sputtered out.

Earlier this month, those hopes were suddenly rekindled when the European Union lifted sanctions against Yugoslavia on 9 October just hours after Vojislav Kostunica was sworn in as the new president of Yugoslavia. Free at last to cut deals with Western benefactors, Vinča has begun talks with organizations in Germany and elsewhere in Europe. “We now have real hopes of finishing TESLA,” says Marina Sokcic-Kostic, a former Vinča physicist who works at the Forschungszentrum in Karlsruhe, Germany. She was in Belgrade for a conference during the demonstrations earlier this month that culminated in the ouster of former President Slobodan Milosevic. “It was a fantastic feeling, a feeling that you belonged to history,” she says.

Sokcic-Kostic is not the only scientist basking in the glow of Yugoslavia's new day. The promise of a democratic nation has emboldened scientists across Serbia and its partner state, Montenegro, to start laying plans with colleagues around the world to resuscitate dormant collaborations or begin new ones. They have a lot of catching up to do. The Yugoslavian scientific community has withered in the past decade, crippled by a massive brain drain and starved for resources. The losses have been so great that the comeback of Yugoslavian science depends largely on how much assistance Western countries and expatriate scientists provide. “The real question may be not what the people in Yugoslavia are ready to do to improve this situation, but what the rest of the world is ready to do,” says Boris Ivancevic, a specialist on fungi at the Natural History Museum in Belgrade.

Indeed, the daunting tasks ahead have begun to sober up scientists intoxicated by newfound political freedom. But in forging a new path, they may find lessons in countries that separated from Yugoslavia in 1991. Croatia and Slovenia are nurturing potentially vibrant—though underfunded—research establishments in their young nations, whereas Bosnia-Herzegovina has been slow in rebuilding its scientific community 5 years after the end of a devastating civil war (see p. 692).

“It must be remembered that science has always been an important part of Serbia's history,” says Mark Edmond Clark, a fellow at the Council on Foreign Relations in New York City. “Things can only get better from here on out.” Adds Martin Prochnik, who until 1995 ran the U.S. State Department's Office of Science and Technology Cooperation, “The end of the Milosevic regime will mean the end of the isolation from the world scientific community.”

## The dark years

Many scientists here yearn for the pre-Milosevic era, a feeling they call “Yugonostalgia.” During the 35 years that Marshal Tito (born Josip Broz) ruled the country, Yugoslav scientists enjoyed a big advantage over their counterparts in other Eastern European countries. Although funding was modest during the Tito era, by maintaining a delicate diplomatic balance—not getting too chummy with either superpower—Tito afforded Yugoslav scientists the opportunity to work in the United States and Western Europe. “Science was a window to the world, and we had excellent connections abroad,” says Croatian biophysicist Greta Pifat-Mrzljak of the Ruđer Bovšković Institute in Zagreb, Croatia.

For most scientific disciplines, says Ivancevic, the decade following Tito's death in 1980 until Yugoslavia broke apart in the early 1990s was the most prosperous period for Yugoslav science. The country as a whole even presented a united front to the world. “There was rather astonishing internal Yugoslav cooperation in working with U.S. scientists, even among ethnic and national groups who ended up killing each other shortly after the war began,” says Prochnik.

Now, Serbian scientists find it painful to compare their situation with that before the breakup of Yugoslavia. “It's difficult to convey the deepness of the tragedy,” said physicist Milan Kurepa of the University of Belgrade in an interview in August. (Kurepa, a veteran fighter for the freedom of the university, died last week after a long illness.) After sanctions were imposed, inflation spiraled out of control. “The living standard was reduced to mere survival,” says Ivancevic. “The common people suffered because of the sanctions, while Milosevic prospered.” Due to widespread impoverishment and inflation, research support plummeted. In 1990, assistant professors at the University of Belgrade could expect to receive a salary of $500 to$600 per month. Now, they receive about $35. Today, “there is no money even for postage stamps,” says Ivancevic. This year through mid-September, his museum has received 2 million dinars (about$33,000) from the government, nearly 75% of which has gone to salaries—after taxes, about \$40 a month for a curator. Consuming the remainder are routine expenses such as heating and electricity bills. Museum researchers must find nongovernmental support for any research projects, collection trips, or costs of attending conferences abroad. A decade ago, the Institute of Physics in Belgrade subscribed to 132 journals; today, seven or eight. “I haven't seen a new textbook from the West for 4 or 5 years,” said Kurepa. “In science, that makes us illiterate. We have no idea what physicists outside Yugoslavia are doing.”

The poor conditions have driven several hundred thousand Serbs with university degrees abroad over the past 10 years. “Many have started successful careers in the West, where conditions for scientific work will be incomparably better than anything even new, democratic Yugoslavia can hope to offer for a long time,” says Ivancevic. “My generation is really all over the world,” says Sokcic-Kostic. She has a son in 10th grade in Germany and would only return to Yugoslavia after he graduates. Indeed, added Kurepa, “the scientific community in Belgrade has almost ceased to exist.”

Ivković, who as Science went to press was expected to lose his post as Serbia's science minister, disputes those claims. “Despite the sanctions, we have managed to maintain a thriving scientific community,” he said in an interview in August, pointing out that the number of Serbian scientists actually rose from 15,100 in 1990 to 15,910 in 1999, as researchers who emigrated were readily replaced. Ivković also pointed out that Serbia currently has a few dozen joint projects with Russia and China, two countries that didn't agree to U.N. sanctions.

Scientists say, however, that the sanctions had long-term effects. Coming back from trips abroad, says electrical engineer Srbijanca Turajlić, “I smuggled vacuum tubes and other spare parts in my suitcases.” Sanctions hit hard even on a personal level. Kurepa, who did his Ph.D. in nuclear physics at the University of Liverpool in the early 1960s before returning to Belgrade, maintained contacts and friendships over the years—until after the sanctions, when he said his British colleagues suddenly fell silent. “They didn't even ask, ‘Milan, do you need a pencil?’” The sanctions, Turajlić says, “were one of the biggest mistakes the West made. We were not able to live under sanctions, let alone do research.”

The NATO bombing campaign in spring 1999 dealt a further blow, bringing research to a virtual standstill for months. “Is it possible to think seriously about science when one sees tens or hundreds of people injured daily?” asks one Belgrade scientist. Rumors even spread that Vinča was on the NATO target list.

## Repairing the damage

Undoing a decade's worth of damage may take years and require grooming a new generation of scientists to take the place of colleagues who went abroad—and probably aren't coming back. “You can definitely forget about the wave of scientists returning to their homeland,” predicts Arseni Markoff, a molecular biologist at the University of Münster in Germany, who sees parallels between the nearly bloodless transition to democracy in Serbia and what happened in Bulgaria, his native land, 10 years ago. Serbian scientists agree. “We do not expect the return of a significant number of scientists to our country,” says Ivancevic. “But we expect that they will continue to help us by sending us information and establishing contacts.”

Cultural changes are necessary too. “In an increasingly impoverished country, moneymaking, whichever way possible, has become the only goal,” says Ivancevic. “I fear that it will take a long time for this attitude to change.” Indeed, predicts Gligor Tashkovich, executive vice president of AMBO LLC's Trans-Balkan Oil Pipeline project, which will be the largest private infrastructure project in the region when construction starts in 2003, “as the economy gets restructured, scientists will be disproportionately squeezed out of jobs. Things will get worse for scientists before they get better, and the only ameliorating condition will be if [the U.S. National Science Foundation] and its counterparts across Europe take a leadership role at this early stage.”

Like the rest of the Yugoslavian scientific community, Vinča and its 400-strong researchers are undertaking the seemingly paradoxical endeavors of rebuilding old ties and burying the past. It's hoping to start afresh with erstwhile collaborators like Oak Ridge National Laboratory in Tennessee and CERN, the European particle physics laboratory in Geneva, which helped set up TESLA's advisory committee before severing ties with Vinča after sanctions were imposed. Renewed cooperation “might attract some of our scientists who have left the country,” says Nebojsa Neskovic, Vinča's associate director for science. Once the cyclotron is finished—Neskovic estimates that the center can be up and running within 30 months after construction resumes—the first project will hook up the hydrogen ion source to produce a beam of 30-million electron volt protons for use in manufacturing radioisotopes and for radiation research. So far, 15 institutes in eight countries, from Italy to Macedonia, have become members of the center.

The future is less certain for the resumption of close scientific ties between the nations of former Yugoslavia. “If democratization in Yugoslavia results in real—not cosmetic—changes of its former structures and goals, we expect the links with the scientific community of Serbia and Montenegro to be more intensive than they were before 1990,” says Croatia's deputy science minister, Davor Butkovic. Georgi Efremov, president of the Macedonian Academy of Sciences, notes that although there is a “paper contract” for scientific collaboration between his academy and the Serbian Academy of Arts and Sciences, there are currently no research projects or exchange of people or lecturers. But, he says, “with democracy in Serbia, collaboration will be reestablished—this is the general feeling for scientists in Macedonia, and I'm sure for Serbian scientists as well.” Croatian biomedical researcher Krešimir Pavelić, of the Ruđer Bovšković Institute, thinks that scientists “should be the ones who take the first step towards the normalization of relations between Croatia and Yugoslavia.”

There are promising signs that the next generation is ready to mend fences. Marko Popovic, a biology undergraduate at the University of Belgrade, was surprised at how warmly he was received when he attended a 2-week biophysics summer school in Rovinj, Croatia, last month. “When I arrived in Zagreb, I was scared, because a Serb tends to get defensive there,” says Popovic. “But once you start discussing science, political problems go away.” At least that's the hope for scientists trying to rebuild their war-torn countries.

9. EASTERN EUROPE

# Serbian University Law Causes Turmoil

1. Richard Stone*
1. *With reporting by Gillian Sandford in Belgrade.

BELGRADE—Whether the University of Belgrade was ruined or saved in May 1998 depends on one's political views. That was when the regime of Slobodan Milosevic pushed through the Serb parliament the University Act, a law that gave the minister of higher education the right to appoint deans at Serbian universities. The act “put Serbia's universities under political control,” says Boris Ivancevic, a specialist on fungi at the Natural History Museum in Belgrade. “The chief criterion was loyalty to the regime.”

The law also mandated that all professors, who are appointed by the deans, sign contracts. “It abolishes all autonomy of teaching,” says University of Belgrade archaeologist Staša Babić. “There are some really fine scholars who signed” for fear of getting sacked and thus making their families suffer, she says. Babić decided to take a chance and refused to sign: “I only have a dog. We'll survive.” She kept her job.

Milan Kurepa of the University of Belgrade physics department said in an interview in August that 58 electrical engineering professors—about a third of the faculty—were either expelled or forced 290into premature retirement. (Kurepa died last week after a prolonged illness.) Two years ago, Kurepa banded together with a handful of other scientists “to fight against the machinery.” They formed the Association of Yugoslav Scientists and Professors and brought attention to their plight in the opposition press. Out of 4200 or so professors at the University of Belgrade, 300 belong to the association. One-sixth of those who joined the association lost their jobs, Kurepa claimed. Many of their replacements were not qualified, he said. “People have to know the university's completely ruined,” Kurepa says.

Serbian science minister Branislav Ivković, who is expected to lose his post with the change in government, disputes this characterization. Appointments are “decided by all the staff of the faculty,” he says. (The act stipulates only that deans may consult teaching staff.) “Simply said, the university structure is now quite similar to [that of] a state university in the U.S.” Ivković argues that “the message of the law is that politics is banned at the universities”: It purged professors who were using their positions and computers—state property—“against the country and against the president.”

Gazo Knezevic, a member of the cabinet of president Vojislav Kostunica, says that the new government will revoke the act. “We expect that the universities in Serbia will transform and modernize,” says Srbijanka Turajlić, an electrical engineer who now chairs the Alternative Academic Educational Network, a shadow university in an office building in downtown Belgrade supported by international philanthropies. The change in government has already brought changes at the University of Belgrade: Rector Jagos Puric handed his resignation to the Serbian government on 16 October, and Ivancevic says as many as 23 of 30 deans at the university have so far resigned. But, says Turajlić, full recovery “will take time.”

10. EASTERN EUROPE

# Science Survives in Breakaway States

1. Robert Koenig

LJUBLJANA AND ZAGREB—As Serb scientists start rebuilding a research system devastated by wars and international ostracism (see main text), they might draw some lessons—and hopes—from colleagues in the nations that won their independence from Yugoslavia during the past decade.

Slovenia, a country the size of New Jersey with 2 million people, has emerged from the chaos of the Yugoslav breakup with the biggest head start in science. One reason is that it escaped unscathed from its 1-week war of independence in 1991. But Slovenia has also tried hard to keep talented young scientists at home, and Slovenian researchers were able to join the European Union's (E.U.'s) flagship research program last year. “Because of our determination to join the E.U. and to focus our priorities on European programs and initiatives, science in Slovenia has improved at a faster rate than in other nations of the former Yugoslavia,” contends plant biologist Lojze Marinč;ek, Slovenia's science minister. In the initial rounds of the E.U.'s Fifth Framework Programme, the ministry says, researchers in Slovenia landed a role in 46 projects, about half as many as did scientists in Poland—a country with a population nearly 20 times as large.

But Slovenian scientists, although proud of the quality of their research, are worried, because government science spending has eroded over the past few years, forcing institutes to channel a higher proportion of their budgets into salary payments. This month, Slovenia's Association of Research Institutes even resorted to newspaper ads to call attention to their plight. Vito Turk, a biochemist who heads Slovenia's biggest research center, the Jožef Stefan Institute, says, “Farsighted initiatives in the past helped Slovenia preserve the quality of its science. But we won't be able to keep up this momentum without more resources for research.”

Slovenia has slowed down the brain drain with a fellowship program called “2000 Young Researchers,” but leading scientists fear that if government and industry don't focus more on R&D, many young researchers will grow pessimistic about their prospects for productive careers in Slovenia. “Without the Young Researchers program, we would have been devastated,” says physicist Robert Blinc. “But what do we do with all these talented young researchers?”

## Shaky start

Like Slovenia, Croatia also has a proud scientific tradition and has for centuries felt closer to its Alpine and Adriatic neighbors than to the Balkan nations to the east. But Croatia paid a much higher price than Slovenia paid for its freedom: a bloody war for independence that took nearly 4 years to settle. Serb rockets killed students and hit university buildings in downtown Zagreb in 1995 and nearly destroyed the university in Osijek, recalls Branko Jeren, rector of the University of Zagreb. The war, and Croatia's involvement in the related Bosnian conflict, sapped the country's economic vitality. And Western European governments barred Croatia from E.U. research programs because of a virulent strain of nationalism nurtured by former President Franjo Tudjman's regime, which ended after Tudjman's death late last year.

With the rise of a more democratic government in Croatia this spring, the nation's scientists are once again eligible to apply with partners for E.U. research programs. And, even though research spending has been low for years, science minister Hrvoje Kraljević told Science that the new government “is committed to increasing support for research.” He is trying to convince Parliament to set aside for R&D a small percentage of the income from the state lottery and from privatization of state-owned companies. That income, he says, could go to setting up something that Croatia doesn't have at the moment: a competitive granting agency. The government now spends only 0.4% of its gross domestic product on science, and “almost nothing comes from the private sector now,” says Kraljević.

Although researchers at Croatia's institutes and universities are hurting because of the budget crunch, many are optimistic that—with more government support and eligibility for E.U. research programs—they will be able to improve the quality of science. “Croatian scientists have had to cope with aging buildings, old equipment, and a lack of innovative technology,” says Krešimir Pavelić, who directs the Ruđ;er Bošković Institute's molecular medicine division. “But there is great potential here because of the young talent.” Pavelić has tapped that potential by aggressively expanding his research division in recent years, and he has even attracted some scientists from abroad.

Nikola Zovko, a physicist who directs the Bošković Institute—Croatia's largest—also sees great potential for growth in Croatian science, but he says scientists have to find more international sources of funding. One possibility is to cash in on one of Croatia's assets: its natural beauty, a long Adriatic coast dotted by historic cities such as Dubrovnik and Split. Croatia now holds about 20 international scientific conferences a year. Such events, says Zovko, “show that good science is taking place in Croatia.”

The problems of Croatian researchers pale in comparison to those in neighboring Bosnia-Herzegovina, governed by a fragile coalition of Muslim, Croat, and Serb ethnic groups. Top scientists fled Bosnia during the war, and few have returned. Some left because of bloodshed, others because their workplaces were destroyed. Energoinvest, once the former Yugoslavia's largest exporter, in its heyday had 11 R&D centers throughout the country. Ten were mostly obliterated during the war, driving three-quarters of the company's Ph.D. researchers abroad, says Bozidar Matic, president of the Bosnia-Herzegovina Academy of Sciences, which itself is estranged from a separatist Bosnian Serb academy. The astronomy observatory outside Sarajevo was also damaged beyond repair by the constant shelling of the war years, and other institutes—such as the Center for Balkan Research, a Sarajevo-based archaeological institute—are wasting away. The center wasn't bombed during the war. Rather, its best scientists simply died or left. Says Matic: “There was a process of ‘negative selection’ in science: The best researchers left.”

11. EASTERN EUROPE

# Adriatic Nations Team Up to Explore Spreading Marine Mystery

1. Robert Koenig

ROVINJ, CROATIA—Climbing aboard what was once a Croatian navy minesweeper, marine chemist Nedad Smodlaka fishes a key from his pocket and unlocks the door to his wet lab. Inside sit stacked rows of seawater-filled bottles. Smodlaka hopes the samples will help to unlock the mystery of a slimy biomass that sometimes despoils the sparkling blue-green waters of the Adriatic Sea. In the process, he also hopes to push open the door to greater international scientific collaboration in the troubled region.

The mystery that Smodlaka is probing is an organic “mucilage” that Croatians call “cvjetanje mora” and Italians “mare sporco.” The biomass, thought to be generated by phytoplankton, entraps plankton and other small sea creatures that produce gases, causing the slime to rise like clouds to the sea's surface. It appears as a translucent white cloud in the sea's depths but turns into a murky mess when it reaches the surface and is pushed ashore by the wind and tides. The mucilage, which occurs only in certain warm summer months, was first reported in 1729. But its appearances, once rare, have become increasingly frequent since 1988, leading scientists to suspect pollution as a major factor. “This is a natural phenomenon,” Smodlaka observes, “but its recent frequency indicates that there are environmental stresses.”

It is not the only sign of environmental stress in the northern Adriatic, but it is perhaps the most visible. And its growing notoriety could lead to a major international program to study pollution and other environmental trends in the Adriatic.

Although researchers have long been interested in the phenomenon, Smodlaka and his team of 35 scientists at the Center for Marine and Environmental Research in Rovinj, on the sea's northeastern coast, haven't always had the resources they need. The turning point came in 1997, when a mass of the slime covered a beach near the vacation residence of Croatia's former strongman leader, General Franco Tudjman. There have been five acute episodes since 1988, says Smodlaka, including a visitation this past summer that has been studied extensively.

With its 700 islands and stunning 5700-kilometer-long coastline, Croatia is particularly sensitive to problems in the Adriatic. But any solution will require the cooperation of several other countries—including Italy, whose Po River is the biggest source of pollution to the northern Adriatic. “Protecting, restoring, and sustaining the Adriatic ecosystem requires the close cooperation of all the coastal countries,” says Slovenian biologist Alenka Malej, who heads the Piran marine station of Slovenia's National Biology Institute. Serena Fonda-Umani, an Italian researcher at the University of Trieste's marine biology laboratory who works closely with Croatian and Slovenian counterparts, notes that “the sea has no borders, and what happens on one side of the Adriatic strongly influences what happens on the other side. If we are not able to work together, we'll never reach a complete understanding of ‘our’ sea.”

Toward that end, researchers from the three countries are examining factors that may contribute to the mucilage. They include the role of bacterial plankton and the hydrochemical dynamics of the Adriatic. The work is part of a widening research effort into problems facing the shallow northern Adriatic, where pollution from the Po and other rivers has fueled the growth of phytoplankton and the depletion of oxygen along the bottom. That anoxia, in turn, has damaged the sea's mollusk and crustacean populations. In a series of workshops in Piran, Rovinj, and Trieste, Slovenia's Malej says that marine scientists from the three nations “were able to identify priority environmental issues and observation needs in the northern Adriatic.”

The next step is to enlist colleagues in an ambitious effort—called the Coordinated Adriatic Observing System—to monitor pollution, currents, and other seawater trends. The joint project is an attempt to share marine data collected by countries that border the Adriatic, including Slovenia, Croatia, Italy, Bosnia-Herzegovina, Macedonia, and Albania. Smodlaka says that the U.S. government has offered to help fund the initiative, and the European Union (E.U.) is weighing a proposal. One problem, Malej says, is the checkerboard political landscape: Italy is part of the European Union and Slovenia has applied for admission, but Croatia and the Balkan nations south of it are outside the club. So getting E.U. funding is tricky. Even so, all the scientists agree that tackling common problems like sea pollution is one good way to increase research cooperation in the region.

Fortunately, Adriatic research has a long, multinational tradition. The first marine station on the sea was founded in 1876 in Trieste, followed in 1891 by a research center in Rovinj opened by the Berlin Academy of Sciences. The German-built facility has survived as an aquarium and research center despite a succession of political masters that included the Austro-Hungarian Empire, Italy, a “quisling” Nazi-era government, and Yugoslavia before the breakup that ended with Croatia's independence.

That history has created a wealth of long-term data about water quality, biodiversity, and weather in the northern Adriatic. Researchers in Rovinj have taken readings of oxygen levels in the region's seawater since the 1920s, for example, and for the past 2 decades the Croatians have teamed up with Slovenian and Italian researchers—as well as U.S. and other European oceanographers—to study the physical, chemical, and biological changes in the Adriatic. A 1996 project involving Smodlaka, Malej, and Thomas C. Malone of the University of Maryland's Horn Point Laboratory in Cambridge, Maryland, concluded that solving the Adriatic's problems will be even tougher than tackling those of the Chesapeake Bay in eastern Maryland because of the “complex mix of sovereign nations representing a diversity of cultures.”

Pointing to the cracked walls and water-blistered plaster at his institute—a building so old that it hosted famous German bacteriologist Robert Koch a century ago—Smodlaka says that the problem is simple. “There is a lot of international collaboration, but not much money.” Even so, he is optimistic that solid research will eventually lead to better equipped labs. “Our building may be falling apart,” he says, “but we're holding together.”

12. ENDOCRINE DISRUPTERS

# Panel Cautiously Confirms Low-Dose Effects

1. Jocelyn Kaiser*

Expert panel says estrogenic chemicals can cause biological effects at levels below those normally found safe, but the implications for human health are unclear

Of all the evidence suggesting that hormonelike chemicals in the environment might be bending the gender of wildlife and people, a lab study 3 years ago sparked the greatest alarm. It suggested that minuscule amounts of a chemical that everyone is exposed to—in baby bottles and Tupperware, for example—could alter the reproductive organs of a developing mouse. Virtually no one had reported effects from so-called endocrine disrupters at such low levels before. The results suggested that the Environmental Protection Agency's (EPA's) plans to begin screening chemicals for estrogenic activity might have to be revised to detect effects at lower levels. The agency's conundrum only deepened when other labs—funded mainly by industry—couldn't replicate the findings.

Faced with the dueling studies and aware of other new research that might resolve the question, EPA enlisted the help of an expert panel.* Earlier this month it met to conduct an extensive review of the data. Chaired by toxicologist George Lucier, who recently retired from the National Institute of Environmental Health Sciences (NIEHS), the 36-member panel concluded that doses of some endocrine disrupters far below those normally found to be safe can indeed cause biological effects in lab animals—a finding that runs counter to the conventional wisdom in toxicology. Panel members cautioned, however, that scientists still don't know what relevance the subtle developmental changes observed in animals might have for human health.

The work that triggered this inquiry was conducted by reproductive biologist Fred vom Saal's group at the University of Missouri, Columbia. Published in 1997, the team's two studies found that at levels below the normal testing threshold, the potent chemical diethylstilbestrol (DES) had peculiar effects on the male offspring of pregnant mice. At some very low doses, it enlarged a fetal mouse's prostate, but levels above this had the opposite effect. If low doses of other chemicals produced effects with this same humped curve, then tests conducted at higher doses might be missing biological effects, vom Saal suggested. But it was the second study that proved most explosive: It found that the plastics ingredient bisphenol A enlarged the prostate in mouse fetuses at the extremely low levels—parts per billion—to which people are typically exposed.

Alarmed about what the results might mean for the multibillion-dollar plastics industry, companies launched a wave of new research on bisphenol A. But attempts by John Ashby at the Zeneca Central Toxicology Laboratory in Cheshire, United Kingdom, among others, to replicate vom Saal's study with the same mouse strain, CF1, failed to find any bisphenol A or DES effects.

The task of sorting out these data fell to the expert panel, formed by NIEHS's National Toxicology Program at EPA's request. Its examination went “far, far beyond the usual peer review,” notes Lucier. Authors of 38 studies submitted their raw data, including unpublished results, to the panel, which also heard presentations from key scientists.

The bisphenol A studies proved to be the biggest puzzle. The negative data included two major new studies by Japanese and industry-supported U.S. groups that found no effects of bisphenol A on developing rats or their offspring. But curiously, when the panel picked apart the protocols and data, a summary statement says it found that both vom Saal's studies and the negative ones were “credible and sound.”

The discrepancies may be attributable to lab animals' exquisite sensitivity to estrogenic chemicals. As scientists noted, many factors—such as whether animals are fed or injected with chemicals, as well as their diet—can influence hormone levels. Another issue, reported in Science last year (20 August 1999, pp. 1190 and 1259), is that rodent strains can vary dramatically in their response to estrogenic compounds. Some scientists speculate that Ashby's CF1 mice might be genetically less sensitive to estrogens than the vom Saal mouse colony, which was maintained as an inbred group for many years.

Not all panelists were satisfied that the discrepancies have been explained, however. Vom Saal's results have “not been reproducibly established,” insists biochemist George Stancel of the University of Texas, Houston, Health Sciences Center. Although a few other groups report that they've replicated the study, they either failed to submit their raw data or panelists found problems with the experiments. And vom Saal has destroyed his original strain of mice—so as to avoid mixing with a new type, he says—so nobody will ever be able to repeat his original bisphenol A experiment. Vom Saal's team reported at the meeting, however, that it has now gotten the same results with the new strain, CD1.

Other studies that showed an effect at low doses, though, were more straightforward, prompting the panel to conclude that low-dose effects from estrogenic compounds “have been clearly demonstrated.” NIEHS researcher Retha Newbold, for example, has replicated vom Saal's inverted U curve for DES when looking at uterine changes in CD1 mice. And preliminary results from other federal studies have found alterations in brain and immune system development in rats from low doses of the plant estrogen genestein and nonylphenol, a chemical in detergents. How these results may relate to disease late in life in animals, let alone humans, is uncertain, say panel organizers. “It's not clear what the biological significance” may be, says pediatrician Lynn Goldman of Johns Hopkins University in Baltimore, Maryland, a former EPA official and a meeting organizer.

Pointing to these uncertainties, EPA's Anthony Maciorowski says it's “premature” to comment until the final report is out next spring. But he suggests that the agency may hold off on modifying its congressionally mandated screening program, which is set to begin testing commercial products in 2003. “If low-dose [effects] appear to play out, they can still be accommodated later,” Maciorowski says. Lucier, for one, believes the evidence is strong enough: “I think there is reason for EPA to revisit” its testing requirements already, he says. But, he adds, the problem is not urgent. “If they're missing something [in the meantime], they're not missing anything catastrophic.”

• *Endocrine Disruptors Low Dose Peer Review, Research Triangle Park, North Carolina, 10–12 October.

13. CLIMATOLOGY

# Does a Climate Clock Get a Noisy Boost?

1. Richard A. Kerr

Ever-present climate noise may amplify a periodic nudging of the climate system, with dramatic past effects and uncertain future implications

Noise has long exasperated researchers searching for signs of subtle regularity in geologic climate records. But noise may soon be viewed not just as an obstacle to understanding but also as an essential driver of climate change over the millennia. Three researchers pondering what could be behind a roughly 1500-year cycle of warming and cooling that most recently gave the world the Little Ice Age (Science, 25 June 1999, p. 2069) are suggesting that this millennial cycle is a combination of two climate drivers, each too weak to have a large effect on its own. When a strictly periodic cycle teams up with just the right amount of thoroughly irregular noise, the combination could achieve “stochastic resonance” and set off the dramatic climate shifts of the last ice age—or the next, possibly human-triggered, Little Ice Age.

In talks at recent gatherings, glaciologist Richard Alley of Pennsylvania State University, University Park; geophysicist Sidhar Anandakrishnan of the University of Alabama, Tuscaloosa; and physicist Peter Jung of Ohio University in Athens have presented an analysis of climate records preserved in Greenland ice cores that suggests stochastic resonance is behind the 1500-year climate cycle. If so, “noise also matters,” says Alley. “What they describe is certainly consistent with stochastic resonance,” says geophysicist Bruce Bills of Scripps Institution of Oceanography in La Jolla, California. “I came away feeling it was a quite plausible case.” No one is claiming the case is proven, but if true it implies that “prediction is going to be really difficult,” says Bills. That would complicate things as humans add in their own “climate noise” over coming centuries.

This isn't the first time researchers have invoked stochastic resonance to explain a climate cycle. In fact, the concept owes its origin to such a cycle. Now applied to everything from signal detection in electronic circuits to the neurophysiology of crayfish, stochastic resonance debuted in the early 1980s as a proposed explanation for the comings and goings of the ice ages every 100,000 years or so. If given a big enough push, the thinking went, Earth's climate system could flip between its two modes of operation, cold glacial periods and warm interglacial periods. But the only obvious means of setting the observed 100,000-year pacing—the rhythmic elongation of Earth's orbit—is too weak a driver to make the switch by itself.

Random—that is, stochastic—noise in the form of short-term climate fluctuations might help, the argument went, if it were strong enough. Increased just to the point at which noise plus the periodic signal suffice to switch the climate system from one mode to the next, noise would “resonate” with periodicity and an ice age would begin most but not every 100,000 years. Sometimes the system would skip a beat or two if by chance the noise were not strong enough when the periodic push was at its height.

Alley and his colleagues reunited stochastic resonance and climate after Alley happened to speak at Ohio University, where Jung applies stochastic resonance to neurophysiology. Alley was looking for drivers for the 1500-year cycle of moderate cooling and warming that punctuates the more dramatic 100,000-year ice age cycle. To test the 1500-year cycle for resonance, Alley, Anandakrishnan, and Jung gauged the durations of millennial climate cycles as recorded by the oxygen isotope composition of the 110,000-year Greenland Ice Core Project (GRIP) ice core from central Greenland. If the cycle were strictly periodic, durations would cluster around 1500 years alone.

If the ice-core isotope signal were entirely noise, there would be no preference for any particular period. But if stochastic resonance were at work, most intervals would fall near 1500 years, far fewer near two cycles or 3000 years, and fewer still near three cycles or 4500 years, with few falling at any fraction of a cycle length. That's what the researchers found in the GRIP core as well as in the nearby Greenland Ice Sheet Project 2 core. The separate peaks of decreasing amplitude with increasing interval length are there but they're very ragged, says Jung. “The problem is we only have a small number of cycles” in the record, says Jung. “You're dealing with small numbers and large fluctuations” of climate.

No one is yet claiming that stochastic resonance is the answer—it hasn't yet solved the ice age problem either—but it's being welcomed as a promising option. “It makes intuitive sense,” says paleoclimatologist James White of the University of Colorado, Boulder. Biophysicist Frank Moss of the University of Missouri, St. Louis, who has demonstrated that stochastic resonance helps paddlefish find food, considers the argument “very convincing. The authors have been careful in applying the statistics.”

If stochastic resonance is indeed operating, says Alley, “I really and truly don't know” what the weak periodic signal is. Suggestions have included periodic flip-flops in deep-sea circulation, solar variations, long-term tidal variations, and orbitally driven processes in the tropical Pacific. The noise may come from the largely erratic behavior of ice sheets, which have been implicated in some of the larger glacial climate oscillations (Science, 6 January 1995, p. 27).

What stochastic resonance would mean for the future is even less clear. When it's involved, “a really hard kick to the system will change the climate,” says Alley. “The interesting question is, can humans kick it hard enough?” Greenhouse warming, ironically enough, could in theory so suppress deep-sea circulation and the warmth that it brings to the North Atlantic that another Little Ice Age could set in, at least around the North Atlantic region—a chilling punishment for making a little noise.

14. NEUROSCIENCE

# Neuroscience Meeting Draws Crowds, Gripes, Loyalty

1. Laura Helmuth

Preparing for their 30th conference, neuroscientists contemplate how the field—and the annual meeting—has grown

The Society for Neuroscience (SFN) was formed in 1969 in part because people were sick of Atlantic City. The New Jersey gambling and boardwalk town was one of the only places in the United States with enough hotel rooms to accommodate the 5000 attendees of the Federation of American Societies for Experimental Biology (FASEB) meeting, and people were starting to feel crowded. “It's one of the more mundane things,” says Lawrence Kruger, chair of the society's history of neuroscience committee, but such a swarm of biologists meant that it was a pain in the neck to get a hotel room, and you had to wait 2 hours to eat dinner. It was enough to make the neurophysiologists in the meeting decide to split off and look for someplace a little more private.

Today, with almost 30,000 members and an annual meeting that routinely packs in about 25,000 people, Kruger hears the same complaints. As the field of neuroscience has exploded, so has the society's meeting, and now it, too, can only be housed in a few cities. But he and other neuroscientists aren't about to fragment their annual meeting, which this year runs from 4–9 November in New Orleans. “Everybody complains about the size, and there are lots of speed bumps and glitches,” concedes SFN president Dennis Choi of Washington University in St. Louis. But because of the meeting's breadth and depth, says Ron Mangun of Duke University, “it's where neuroscientists go.”

Why do so many researchers come back every year, knowing they'll have to stand in long lines, lug around 25 kilograms of program and abstract books, and squint at slides from the back of a room that seats 4000? Some of the reasons neuroscientists cite apply to any scientific conference: the thrill of hearing about hot new research and the chance to catch up with friends.

But others say that the best reason to go to the SFN meeting every year is precisely because it's so huge and integrates the ever-growing number of subdisciplines in the field. It's a place to pick up inspiration or technical advice from people in related—or not-so-related—fields, get feedback from a wide range of perspectives, and forge interdisciplinary collaborations. “We're all linked by an interest in understanding the brain in a larger sense,” says Choi, “whether at the genetic or the system level.”

In 1970, 1 year after it had been founded, the society counted 1100 members. Many were neurophysiologists, experimental psychologists, or psychiatrists, says past SFN president and founding member Edward G. Jones of the University of California, Davis. It was an exciting time, he says, as new staining techniques were enabling neuroanatomists to track where axons went in the brain and to determine which neurons communicated with each other. It was also becoming technically feasible to record from electrodes in the brains of animals while they performed lab tasks. And in neurochemistry, researchers were starting to analyze how drugs interact with receptors on nerve cells.

As neuroscientists took up new tools and researchers in other fields turned their attention to neurons, new subdisciplines joined the society. Neuroscience “keeps drawing other fields into itself,” explains Jones. Invertebrate neurobiologists started joining the society in large numbers a few years after it had been established, bringing tools for studying the development of the nervous system, particularly in such model organisms as Drosophila and Caenorhabditis elegans. Studies in the early 1970s, recalls Jones, revealed the astonishing commonalities among the nervous systems of different animals at the earliest stages of development. Cellular and molecular neurobiology took off in the late 1970s and early 1980s, thanks in part to sophisticated electron microscopy that brought into view the actual synapses and other components of the nerve cell.

Some of the latest tools stem from the rapid-fire advances in molecular biology and genetics over the past 15 years, says Steven Hyman, director of the National Institute of Mental Health. These are “allowing us to ask questions that weren't even vaguely possible” when the society was founded, he says. For instance, some of the most intriguing mouse mutants, formed by traditional knockout or transgene techniques, have behavioral consequences. And that enables researchers to map genes to biochemical networks, brain circuitry, and complex behavior. Tools such as bioinformatics for making sense of that outpouring of genetic data, he adds, have only been available for about 5 years.

Brain-imaging techniques have also improved rapidly in the past 10 years, particularly with the adoption of functional magnetic resonance imagery and multielectrode arrays that enable researchers to record from many nerve cells at once. With such techniques in hand, says Hyman, researchers can ask “fundamental questions about cognition and emotion: how the brain focuses attention, how working memory works … which circuitry is used by fear in the human brain.”

During the 1990s, the congressionally declared “decade of the brain,” research on the neurobiology of disease really took off, says Choi. Prior to that time, neuroscience was “largely devoid of practical applications,” he says, but now it has become clear that the field has “imminent applications for medicine, psychiatry, and neurosurgery.” Jones agrees. Understanding the molecular basis of neural development, he says, might uncover treatments for spinal cord injuries. Research on the molecular basis of neuropsychiatric diseases such as schizophrenia could lead to better pharmaceuticals, and stem cell research might provide options for reversing degenerative diseases.

The ability to ask fundamental questions that once seemed impenetrable has attracted a lot of new researchers to the field, says Kruger. “There's so much to be done, and the tools are here. You put a shovel into the ground, and you dig up gold.” And new researchers—who might be a bit more aware of all the dirt surrounding the gold—are flooding into the society. Indeed, of almost 29,000 members, about 20% are students, and the annual meeting is often their first chance to present their work before a broad audience.

“I enjoy watching grad students get wide-eyed,” says Paul Letourneau of the University of Minnesota, Twin Cities. He says new researchers are either completely energized by the conference or totally devastated—because they realize there's so much they'll never know, or because the research question they thought was so esoteric is actually being explored by competitors from around the globe. As Kruger points out, even established researchers complain that, since the field has grown so enormously, people go to the SFN meeting and think, “No matter what I do, no matter what I present, it seems like such a trivial increment in the face of this astonishing advance.”

These astonishing advances are reported in upward of 12,000 abstracts sorted into 874 sessions during the 6-day meeting. All that makes for one huge conference, says Mangun. Because of all the distractions, “it's very common to hear, ‘Oh, I hate that meeting,’” he says. Yet people come, year after year, even those who, like Mangun, have helped found other, more focused conferences. Mangun compares the SFN meeting to New York City. “Some people hate the crowds and the hassle, but where else can you go to the best Russian restaurant one night and the best Thai the next?”

15. NEUROSCIENCE

# Tips for Neuroscience Neophytes

1. Laura Helmuth

Writing an abstract for the Society for Neuroscience (SFN) meeting is an art. Submissions are due in late April or early May for the November meeting, says SFN program committee chair Virginia Lee, because it takes half a year to process the more than 12,000 submissions. So researchers have to write abstracts precise enough to land them in the appropriate session and attract people to their presentation, yet open-ended enough to cover fresh data come conference time.

No matter how vague the abstract, however, it won't be rejected. Every member of the society is allowed to submit one abstract per year, and every one is accepted. This practice helps to maintain quality, SFN president Dennis Choi says, as most people present their best research, notwithstanding the occasional joke poster, such as one a few years ago exploring pain mechanisms in Mr. Potato Head.

Should scientists present their research as a talk or a poster? A few years ago, the answer would have been to aim for a slide session, which was considered more prestigious. But that has changed for the SFN meeting, says Choi, as scientists from some of the best labs opt to give poster presentations instead. As Choi explains, “you give away too much to sit in a slide session for a whole half-day” when you could be out prowling the poster aisles and talking to people.

Students preparing for their first professional meeting, or those who could stand a quick review of the basics, can read a how-to guide written by Beth Fischer and Michael Zigmond of the University of Pittsburgh at www.pitt.edu/∼survival/attend.html. The manual, which grew out of advice given to their own students preparing for the overwhelming SFN meeting, covers everything from what font size to format different parts of a poster to how to dress and which kinds of sessions to attend.

National Institute of Mental Health director Steven Hyman has some advice as well: New conference-goers should attend “some of the symposia not in their area to learn about the wonderful breadth of neuroscience.” The opportunity to learn what's hot in many different fields, he says, “is the glory of this meeting.”

16. # Synapses Call the Shots

1. Marcia Barinaga

Neuroscience research is shedding light on how neurons delegate their protein synthesis, shipping some messenger RNAs out to the dendrites, where they are translated to protein under the control of local synapses

“Dendrites are the brains of the neurons,” neuroscientist Jim Eberwine of the University of Pennsylvania in Philadelphia likes to say. And with good reason. These highly branched structures do the neurons' computations, receiving and adding up signals coming in from other neurons through contacts called synapses. But dendrites do far more than just tally those signals; they also respond to them by strengthening or weakening their synapses. For example, repeated signals to a particular synapse may strengthen that synapse, making it respond more strongly to future signals; other activity patterns can weaken the synapse's future responses. These changes are believed to be at the very root of learning and memory, enabling the brain to learn from and adapt to the information it receives and to store the memories for future use.

These synapse changes require a supply of new proteins, and those proteins must be directed to specific synapses, because synapses undergo changes independently of their neighbors. That presents neurons with a distribution challenge, because the dendrites form a finely branched maze whose synapses may be hundreds of micrometers from the neuron's cell body. Now, after decades of piecing together hints, neuroscientists are building a detailed picture of how this made-to-order protein synthesis occurs. The emerging picture is one in which synapses have “some kind of autonomy,” says neuroscientist Kelsey Martin of the University of California, Los Angeles (UCLA). “You are regulating gene expression very locally within a single cell.” Indeed, it appears that finely controlled, decentralized protein production may contribute to learning and memory in a way that neuroscientists could scarcely have imagined 20 years ago.

That's because the new findings provide a fresh view of the balance of power within cells. Biologists have traditionally viewed the nucleus and its surroundings as the command center of the cell, with the nucleus supplying messenger RNAs (mRNAs) to the ribosomes, tiny protein factories that wait nearby to translate the mRNAs into proteins. Although that general scheme operates in all cells, it appears that neurons have adopted an additional strategy for supplying proteins to their distant dendrites: mRNAs are marked for delivery to the dendrites, where ribosomes function as protein-production outposts; both transport and production are under the control of nearby synapses.

## First inklings

This revolution in thinking hasn't come overnight. For nearly 2 decades researchers have had evidence that proteins may be made in the dendrites. But recent experiments are fleshing out the fine points, showing how neurons mark mRNAs for transport to the dendrites from the nucleus, and how the activity level of synapses in the dendrites regulates that transport and the mRNAs' eventual translation.

The excitement over protein synthesis in the dendrites began in 1982, when neuroscientist Oswald Steward, then at the University of Virginia in Charlottesville, noticed that electron micrographs (EMs) of dendrites showed ribosomes apparently linked to mRNAs, which suggested that the dendrites were making protein. What's more, says Steward, now at the University of California, Irvine, “the localization of the ribosomes was selective.” They weren't just anywhere in the branching dendritic tree, they were in the dendritic spines, the fine, twiglike projections that form synapses. The ribosomes seemed perfectly positioned to supply the proteins synapses need to make the long-term changes that are key to learning and memory. “He showed definitely that the protein-synthetic machinery was there,” says Brown University neuroscientist Justin Fallon. And that provided “a gorgeous mechanism” for regulating the production of synapse-specific proteins, adds neurobiologist Mark Mayford of the Scripps Research Institute in La Jolla, California.

The EMs certainly suggested that proteins were being made in the dendrites, and other researchers verified by immunostaining that all the necessary components of the protein-producing machinery were there. But it wasn't until 1996 that researchers got direct evidence for that protein synthesis. Eberwine and postdoc Peter Crino did this by severing the dendrites of cultured neurons and sucking the cell bodies out of the culture dishes, leaving the disembodied dendrites behind. They then injected mRNAs into the dendrites and showed that those mRNAs were translated into protein.

## The missing link

But many questions remained. Perhaps most important, researchers had not shown a link between dendritic protein synthesis and the synapse changes associated with learning. Among the first groups to forge that link was Erin Schuman's at the California Institute of Technology (Caltech) in Pasadena. The team had shown that a protein called brain-derived neurotrophic factor (BDNF) strengthens synapses in the hippocampus, a brain area associated with some kinds of learning. The strengthening requires fresh proteins, and to learn whether those proteins are made in the dendrites or the cell body, the team took advantage of a feature of hippocampal anatomy: The cell bodies of certain neurons are in one layer of the hippocampus, but their dendrites are in another layer, some distance away. In experiments reported in 1996 the researchers simply sliced the layers apart in cultured hippocampal tissues, disconnecting the dendrites from their cell bodies. When they added BDNF to the cultures, it still strengthened the synapses, which meant, Schuman says, that the disconnected dendrites “had to be the site of protein synthesis.”

Experiments like Schuman's did not reveal the identity of the proteins made in the dendrites, but other researchers were working on that question. One protein they found is calcium-calmodulin-dependent kinase II (CaMKII), an enzyme that adds phosphate groups to other proteins. CaMKII is a hot research topic because it plays a key role in the synapse strengthening linked to learning and memory; when it turned out to be translated in the dendrites, it became a model protein for studying that process as well. Last year, Mary Kennedy, postdoc Yannan Ouyang, and their colleagues at Caltech showed that CaMKII is made in the dendrites of neurons whose synapses have been stimulated in a way that triggers long-term potentiation (LTP), a form of synapse strengthening that appears to be involved in certain kinds of learning. Using antibodies to CaMKII, they detected a 30% increase in the amount of the protein in the dendrites after triggering LTP. They were able to block the increase with protein synthesis inhibitors, suggesting the increase was indeed due to new protein synthesis.

Although CaMKII synthesis had already been shown in the dendrites, Kennedy's experiment provided important evidence that synaptic activity could boost that synthesis. What's more, the translation that is boosted had to be local, Kennedy says, because the new CaMKII began to appear in the dendrites within 5 minutes of the stimulation—much too fast for it to have arrived from the cell body. The idea that synapse activity triggers local enzyme synthesis was reinforced in March, when A. J. Scheetz, reporting in Nature Neuroscience on work he did as a postdoc in Martha Constantine-Paton's lab at Yale University, showed that the new CaMKII protein was made when he activated synapses in synaptoneurosomes, which are preparations of intact synapses that have been broken off from their cell bodies.

The Scheetz team's work also suggests how synapse stimulation turns up translation of the CaMKII mRNA, an important issue because this mRNA is always present in high concentrations in the dendrites. Scheetz and Constantine-Paton found that stimulation actually turns down overall dendritic translation by inhibiting a part of the translation machinery called an elongation factor. That makes elongation the rate-limiting step in protein translation and allows proteins whose mRNAs are abundant in the dendrites, like those of CaMKII, to tie up more of the ribosomes and make proportionately more protein. “It is a way of having a pool of transcripts available but not utilized” until they are needed, Scheetz says. Some viruses use a similar strategy, flooding host cells with their own mRNA while repressing production of host proteins.

## A second mechanism revealed

CaMKII had more lessons to divulge, including a second means of translation control operating in the dendrites—a means that neurons share with developing oocytes. Brown's Fallon, working with Joel Richter of the University of Massachusetts Medical School in Worcester, found a code in the tail end of CaMKII mRNA—the so-called 3' untranslated region—that appears to trigger the mRNA's translation.

Richter's lab studies translation in developing oocytes, in which some mRNAs have a sequence in their 3' untranslated region called the cytoplasmic polyadenylation element (CPE). When a protein called CPEB binds to the CPE, that triggers the addition of a string of adenine nucleotides to the RNA's tail (polyadenylation), which in turn activates the mRNAs for translation. Knowing that translation occurs in dendrites, Fallon encouraged Richter to look for CPEB there. The team found it.

Then Fallon and his postdoc Dave Wells joined forces with Richter and his postdoc Lin Wu and found that the CaMKII 3' untranslated region contains a CPE. What's more, when they strengthened synapses in the brains of young rats, CaMKII mRNA became polyadenylated and newly made CaMKII appeared at the synapses. To see if the increased synthesis was due to polyadenylation, they mutated the CPE. That mutation abolished the 3' untranslated region's ability to increase protein translation in cultured neurons, confirming their hypothesis.

Fallon says the elongation factor and polyadenylation schemes for boosting translation are “not at all incompatible.” The phenomenon that Scheetz and his colleagues observe is “all over within 30 minutes,” he says, and that is when his team just begins to see polyadenylation. Modification of the elongation factor may provide the first boost to CaMKII translation, he suggests, and mRNA polyadenylation may take over later.

## Mail order

Regulating translation is just one way that active synapses boost their protein supplies. It appears to be a key role for proteins like CaMKII, whose mRNA is found at high levels in the dendrites whether or not synapses are stimulated. But researchers studying another mRNA uncovered a whole different level of regulation, in which active synapses order up more mRNA to be shipped out to them from the nucleus. This was dramatically demonstrated for activity-regulated cytoskeletal-associated protein (Arc). In 1995, Greg Lyford, Paul Worley, and their colleagues at Johns Hopkins University in Baltimore, Maryland, found that Arc mRNA is quickly made in the nuclei of neurons that have undergone LTP—hence the designation “activity-regulated”—and shows up in the dendrites. Steward joined Worley to see if the mRNA goes specifically to active synapses.

The researchers used a trick to trigger LTP in one of two layers of the hippocampus of rats, depending on where they placed their electrode. After triggering LTP in one layer, they stained slices of the rat's brains for newly synthesized Arc mRNA and found that it concentrated specifically in the layer that had undergone LTP. If they triggered LTP in the other layer, the Arc mRNA went there. At both sites, it was translated into Arc protein. “The really cool thing was that it went to the activated synapses,” says Brown's Fallon. Neuroscientist Kenneth Kosik of Harvard Medical School in Boston agrees: “For RNA to navigate in that [dendritic] tree, say, to take the first left- and then two right-hand turns and get to the right synapse—it is just an extraordinary regulatory process that is going on there.”

## Postal codes

While the mechanism that directs proteins precisely to activated synapses remains a mystery, it is clear that any mRNA that travels to the synapses must carry a sequence that serves as a general “postal code” to direct it out to the dendrites. Several teams have searched for that code. In 1996, Mayford of Scripps, who was then a postdoc in Eric Kandel's lab at Columbia University, decided that “the likeliest place for a signal to be was the 3' untranslated region,” where other regulatory codes have been found.

To test that hunch, he spliced the DNA encoding the 3' untranslated region from the CaMKII mRNA to DNA encoding a bacterial enzyme that turns tissue blue. When he put the hybrid gene into mice, he found blue color in the dendrites of the animals' brains. That meant the 3' untranslated region was sufficient to direct the mRNA to the dendrites. Conversely, as Stephan Miller, a postdoc in Mayford's lab, will report at the upcoming Society for Neuroscience meeting in New Orleans, swapping the CaMKII mRNA's usual 3' untranslated region for another RNA sequence effectively locks the mRNA in the cell body and keeps it from finding its way to the dendrites.

This appears to be a widely used mechanism for directing mRNAs to the dendrites. For instance, last year Stefan Kindler's team at the University of Hamburg in Germany reported that part of the 3' untranslated region of the mRNA for microtubule-associated protein-2 directs that mRNA to the dendrites.

That 3_untranslated region code apparently works by linking mRNAs to microtubules, fibers that form a sort of monorail system for transporting materials around cells. Harvard's Kosik and postdoc Martha Rook recently devised a scheme for watching the movements of specific mRNAs in living neurons. They hooked the 3' untranslated region from the CaMKII mRNA to an RNA sequence that binds a fluorescent protein. As they reported in the September Journal of Neuroscience, they saw the fluorescently labeled RNAs move gradually into the dendrites of cultured neurons with an oscillating motion. When they activated the synapses, the outward movement of the mRNAs was “more directed,” says Kosik. “They seemed to go over a longer distance without movement back.” (See movies of the RNA movements at neuro-oas.mgh.harvard.edu/alzheimers/Rooketal/rook_et_al.html)

Wondering which proteins might escort the dendrite-bound mRNAs on their monorail journey, researchers turned to an RNA-binding protein called Staufen. Staufen was discovered shuttling mRNAs in developing fruit fly embryos, but researchers later found a mammalian form of Staufen that appears in brain neurons. Early last year, Michael Kiebler, Carlos Dotti, and their colleagues at the European Molecular Biology Laboratory in Heidelberg, Germany, reported that Staufen is present both in RNA-containing particles in the cell body and in dendrites of hippocampal neurons from rat brains. Evidence that Staufen might help transport mRNAs to the dendrites of rat neurons came last fall when Kiebler's team, now at the Max Planck Institute in Tübingen, Germany, showed that it travels on microtubules into the dendrites. And Caltech's Schuman reported at last year's Society for Neuroscience meeting that neurons carrying a defective form of Staufen had less mRNA in their dendrites.

Another unresolved issue concerns how mRNAs know which synapses to go to in the vast mazelike network of dendrites. An mRNA heading for a specific synapse must make many choices at forks in its path, and that clearly requires more than just a dendrite-specific shipping label. When the synapse sends its request to the nucleus, says Schuman, “it is possible that some kind of information about the identity of the synapse that sends the signals could also be transported,” leaving a sort of breadcrumb trail that the mRNA can follow back. But no one yet knows what that information might be.

Once the mRNA finds the right neighborhood, several teams have evidence that activated synapses wear tags that direct mRNAs to their doors. UCLA's Martin, working with Andrea Casadio as postdocs in Kandel's lab, got a clue as to the nature of the tags when they showed that synapses from the sea slug Aplysia can mark themselves even when protein synthesis is blocked. That, says Martin, means the tags must not be newly made proteins. Instead, the researchers suggest that the tags may be placed by an existing enzyme, perhaps a kinase that tacks phosphate groups onto the activated synapse.

One big unanswered question is how activated synapses place their orders with the nucleus for more mRNAs. Martin, Casadio, and Kandel found that when they blocked protein synthesis at the synapses, synapse activation no longer triggered long-term synaptic changes. Those changes require gene activation in the nucleus, says Martin, a result that “tells us that local protein synthesis is involved in the feedback” that triggers that gene activation. What's more, Kandel's and Eberwine's teams have found that the dendrites make a variety of proteins known to regulate gene transcription in the nucleus; it is possible that these proteins may be the messengers.

As researchers press on, they are filling in a picture of a complex loop of communication and supply, with synapses acting as local managers wielding more power than anyone might have imagined, to get what they need for carrying out the critical task of storing memories.

17. # Translational Roots for Mental Retardation?

1. Marcia Barinaga

A flurry of findings points to protein translation in the dendrites of neurons as a key feature leading to the changes at synapses that are vital to learning (see main text). And one recent discovery suggests that when this translation goes awry, it can lead to mental retardation. In 1997, William Greenough and Ivan Jeanne Weiler of the University of Illinois, Urbana-Champaign, working with Jim Eberwine of the University of Pennsylvania in Philadelphia, surveyed the messenger RNAs (mRNAs) found in synaptoneurosomes, preparations of intact synapses broken off from their cell bodies. One mRNA turned out to encode the fragile X mental retardation protein (FMRP), whose gene is mutated in fragile X syndrome, the most common inherited form of mental retardation. What's more, the researchers found that this mRNA became linked to ribosomes—an indication that it was being translated into protein—just minutes after they stimulated the synapses. New FMRP protein also appeared after stimulation, confirming that the protein's production is triggered by synaptic activity.

The work may shed some light on how FMRP mutations cause mental retardation. FMRP is an RNA-binding protein, and recent findings in Greenough's lab suggest that FMRP is needed for protein synthesis near synapses in response to stimulation. When his team stimulated synaptoneurosomes from mice that lack the FMRP gene, the stimulation didn't boost translation as it does in synaptoneurosomes from normal mice. “That suggests that the protein plays a positive role in regulation of translation,” Greenough says. But “the real story from the clinical point of view,” he adds, “is going to be [finding out which] proteins are regulated.”