# News this Week

Science  09 Jun 2000:
Vol. 288, Issue 5472, pp. 1714
1. PLANETARY SCIENCE

# Most-Common Meteorites Find a Home Among the Asteroids

1. Richard A. Kerr

Washington, D.C.—New measurements from a pioneering spacecraft may finally put to rest one of the longest running mysteries in planetary science: Where do meteorites come from?

Nobody seriously doubts that the crashing and banging about of the asteroids between Mars and Jupiter yield the bits of rock that fall from our sky. But meteoriticists have spent more than a century dissecting tons of samples down to the nanometer scale, without being able to pin down their origins. The puzzle is that the most common meteorites—so-called ordinary chondrites, thought to be bits of the primordial building blocks of the solar system—don't appear to have come from the most common asteroids, the S-types. Their colors simply don't match. Last week, however, a group of researchers attending the spring meeting of the American Geophysical Union here announced that they have lifted the veil of at least one S-type asteroid—the 31-kilometer-long Eros now being orbited by the NEAR Shoemaker spacecraft—to reveal its true nature. Eros, it appears, is made of the same stuff as ordinary chondrite meteorites.

That conclusion comes from NEAR Shoemaker's first-ever analysis of the elemental composition of an asteroid. “I'm confident we've got an ordinary chondrite; everything's consistent with it,” says NEAR Shoemaker team member Lucy McFadden of the University of Maryland, College Park. Some team members are anxious to see elemental analyses of more parts of Eros before they are willing to declare the meteorite mystery solved, but “you can tell which way it's going,” says team member Scott Murchie of the Applied Physics Laboratory in Laurel, Maryland. The elemental link between one S-type asteroid and ordinary chondrites reinforces recent findings suggesting that long exposure of ordinary chondrite asteroids to the rigors of space has altered their appearance, to the confusion of Earth-based observers.

Ground-based astronomers searching for the source of ordinary chondrites were perplexed because wherever they looked, they saw red. Even though some of the S-type asteroids seemed to have the right mix of minerals to be ordinary chondrites—to judge by the absorption of specific wavelengths in the visible and near-infrared—they all had a subtle reddish tint. “In my dissertation, I thought Eros was a stony-iron [asteroid], a differentiated asteroid,” says McFadden. Like many other planetary scientists, McFadden had concluded that heating and melting had separated Eros's rock and metal, a process called differentiation, producing chunks of metallic iron that would give the reddish tint. But it was obvious to everyone that ordinary chondrites never got heated to the point of melting.

To explain this conundrum, some researchers argued that Eros and other S-type asteroids might instead be undifferentiated ordinary chondrites reddened only on their surface by the “space weathering” of micrometeorite impacts and the solar wind. Others called such wholly theoretical agents of space weathering so much “foo-foo dust.”

The designers of the NEAR Shoemaker mission hoped that their x-ray-gamma ray spectrometer (XGRS) instrument would settle the differentiation versus space weathering debate once and for all. The XGRS measures the distinctive high-energy electromagnetic emissions of specific elements. Color might be altered by space weathering, the thinking went, but elemental composition could not be. A good idea, everyone agreed, but NEAR would have to get in close to record the faint emissions reliably.

At the meeting, XGRS team leader Jacob Trombka of NASA's Goddard Space Flight Center in Greenbelt, Maryland, reported the first XGRS analyses from a distance of 50 kilometers. On 4 May, a major solar flare had bathed Eros with x-rays that stimulated x-ray fluorescence from its magnesium, aluminum, silicon, and iron, most brightly from a patch 6 kilometers across. The XGRS recorded x-ray fluorescence at specific wavelengths characteristic of each element. The result: By elemental composition, Eros falls solidly in the realm of the ordinary chondrites, Trombka reported. Specifically, it resembles the L and LL classes of ordinary chondrites. “At this point, we are indicating that Eros is undifferentiated,” said Trombka, “or partially differentiated.” (A few rare, “partially differentiated” meteorites come from asteroids that just began to melt, separating their rock and metal incompletely.) “The preponderance of evidence is that it's not differentiated,” says meteoriticist Timothy McCoy of the XGRS team and the National Museum of Natural History in Washington, D.C., “but until we have more coverage of the asteroid, it's hard to say for sure.”

McFadden reported corroborating evidence at the meeting from NEAR Shoemaker's multispectral imager system and near-IR spectrograph. Eros's color and therefore mineralogy most closely match that of ordinary chondrites, she said, in particular the same L and LL classes indicated by elemental composition, and its color is nearly uniform across the asteroid, whereas a partially differentiated body might have patchy colors. “We see no evidence for differentiated assemblages” in the spectral data, says McFadden. “We're getting the same answer from the two techniques.”

Some researchers are further convinced that Eros is an ordinary chondrite in disguise, because they think they finally know how ordinary chondrites can mask themselves in a red cloak. At the Lunar and Planetary Science Conference (LPSC) last March in Houston, members of the Lunar Soil Characterization Consortium reported that micrometeorite impacts and the solar wind cause reddening, at least for the only rock exposed to space weathering that they could get their hands on—lunar “soils” returned by Apollo astronauts. Microscopist Lindsay Keller of MVA Inc. in Norcross, Georgia, and consortium colleagues showed how specks of iron less than 10 nanometers in size revealed by transmission electron spectroscopy account for most of the mysterious lunar reddening. “We've identified the culprit behind the space-weathering effect,” said Keller.

On the moon, according to the picture developed by the members of the consortium, micrometeorites and the charged particles of the solar wind release the iron of rock particles. It is reduced to the metallic state and deposited as “nanophase iron” on soil particle surfaces. Planetary scientist Bruce Hapke of the University of Pittsburgh told the LPSC that, according to his calculations, the solar wind alone can create enough nanophase iron to redden asteroids. “His model and the lunar soil results fit together perfectly,” says planetary scientist Carlé Pieters of Brown University and the consortium. “S asteroids fit with what you'd expect for a space-weathered ordinary chondrite.” Meteoriticist Harry McSween of the University of Tennessee, Knoxville, agrees. “I thought space weathering was a rather bizarre idea when I first heard it,” he says, “but I can't see any way around it now. It's exciting that NEAR Shoemaker is orbiting a body that is like the most common type of meteorite that falls to Earth.”

2. PLANT GENETICS

# From Genome to Functional Genomics

1. Jocelyn Kaiser

Plant scientists are an impatient lot. They are about to complete the first genetic sequence of a flowering plant, a wild mustard called Arabidopsis thaliana. But even before the last A's, C's, G's, and T's are deposited in GenBank, a group of plant scientists has hatched an ambitious plan for the next phase: figuring out the function of all 25,000 genes. Announced last week, the plan, which has the blessing of the National Science Foundation (NSF), came with another bit of good news for the Arabidopsis community: the unexpected release of a set of molecular markers for finding those genes.

The 130-million-base-pair Arabidopsis genome is expected to be fully sequenced in July and published by year's end, 3 years ahead of schedule. Already, information gleaned from decoding this simple plant—the equivalent of the lab mouse—has made “a quantitative change” in research, says Carnegie Institution plant scientist Chris Somerville, whittling the time for isolating genes from years to weeks and thus speeding genetic discoveries ranging from more healthful soybean oil to a protein that may lead to faster growing crops.

Not content to rest on their laurels, Arabidopsis experts now want to determine what proteins are expressed by every single gene, each protein's job within the cell, and their biochemistry—a task that could take 10 years and cost $500 million. The 2010 Project, as it's called, was fleshed out at a January workshop at the Salk Institute for Biological Studies in La Jolla, California; it was recently released on the Web (www.arabidopsis.org/workshop1.html) and is also summarized in this month's issue of Plant Physiology. Proponents say the multinational project will shed light on a host of questions—from how gene expression in any species is influenced by environment to the minimum number of genes needed to make a plant. The group's ultimate goal is to create a “virtual plant” on the Internet, where scientists can click on an Arabidopsis cell at any stage of development, from seed to seed-dropping adult, and see every protein being expressed and the connections among them. However, plan co-author Joe Ecker of the University of Pennsylvania in Philadelphia cautions that the 2010 Project will take them only partway there; for now, they will settle for knowing what all the individual proteins do. That alone is an enormous job. The 2010 Project will first support “genome technology centers” that will supply the necessary tools, such as DNA chips for studying gene expression, libraries of cloned genes, and knockout strains. The project is likely to draw on the talents of labs already gearing up to do high-throughput functional genomics of the nematode C. elegans, fruit fly, and human. Firmly behind the proposal, NSF has asked for$25 million for the 2010 Project for fiscal year 2001, an amount that Ecker hopes will grow or be supplemented by other agencies.

Also last week Cereon Genomics LLC, a subsidiary of Pharmacia Corp., released a set of more than 39,000 SNPs, or single-nucleotide polymorphisms, gene hunters' new favorite tool (www.arabidopsis.org/cereon/index.html). Until now, only about 400 SNPs have been publicly available for Arabidopsis. “It's a huge number if you consider the genome size,” says David Meinke, an Arabidopsis researcher at Oklahoma State University in Stillwater—enough to isolate nearly all the genes. What's more, says Somerville, Cereon is releasing the SNPs with virtually “no strings,” as academic and nonprofit users are free to patent discoveries made with these SNPs. With that and a major functional genomics project in the works, Arabidopsis researchers are clearly on a roll.

3. BIOTECHNOLOGY

# Disease Group Invests in Do-It-Yourself Drugs

1. Eliot Marshall

Chafing at the slow pace of commercial drug development, a disease advocacy group set out last week to finance new medicines for its constituency. On 31 May, the Cystic Fibrosis (CF) Foundation of Bethesda, Maryland, announced that it will invest at least $30 million in a small biotech firm, Aurora Biosciences of San Diego, to identify compounds that might prove useful in treating CF. This project, fueled initially by a donation of$20 million from the Bill and Melinda Gates Foundation, marks a new departure in the growing trend of patient groups taking charge of biomedical research.

The plan calls for Aurora to screen several hundred thousand molecules in its library over the next 5 years and identify two or three that might be candidate drugs for CF. If this approach yields some promising “hits,” the CF Foundation plans to pay Aurora an additional $16.9 million to prepare the candidates for clinical trials. Carrying the drugs through to final approval, however, would require coinvestment by a major pharmaceutical company. Profits would be shared among the CF Foundation and its business partners, but the foundation would immediately plow all of its own royalties directly back into research on new therapies. CF Foundation president Robert Beall thinks this new “virtual drug company,” a hybrid profit-nonprofit venture, is unique in the pharmaceutical industry. His group decided to take the plunge into drug R&D because it didn't want to wait for manufacturers of small-molecule drugs to take an interest in CF. A decade ago when the CF gene was discovered, researchers hoped new drugs would follow close behind. The discovery yielded a wealth of information about what goes amiss in the disease, but translating those insights into therapies has been slow. The CF Foundation is involved in at least 20 collaborative projects and is now supporting clinical trials of gene therapies, using three different types of gene transfer vectors. But this is the first time it has tried to lead the discovery process itself. Big drug companies have not been drawn to the field, Beall notes, because the number of CF patients who might buy a drug is relatively small—only about 30,000 in the United States. And he says that “when we tried to get them involved” in searching for interesting new compounds, “they didn't return our calls.” So the foundation hired a consultant to vet innovative small firms; they quickly settled on Aurora. The company maintains a library of 400,000 small molecules that can be screened at high speed for medical applications. Aurora is a particularly good fit for the CF Foundation, says Beall, because it specializes in assaying proteins that permeate the cell membrane, based on a proprietary blue versus green fluorescence test developed by Roger Tsien and colleagues of the University of California, San Diego (Science, 2 January 1998, p. 84). CF is a disease in which chloride flow through the cell membrane is restricted. Aurora will use cells from CF patients to test whether compounds help restore normal ion channel function, says Paul Negulescu, vice president for discovery biology. “We provide the discovery engine,” he says, “and [the CF Foundation] provides an extensive and sophisticated [drug] development network.” The foundation manages a clinical trial network based at eight centers around the country, coordinated by a team at the Children's Hospital of Seattle. This approach, Negulescu says, could serve as the model for “a new type of drug-discovery process” for other orphan diseases, including those that primarily affect poor nations. Francis Collins, director of the National Human Genome Research Institute and co-discoverer of the CF gene, says “this roll-up-your sleeves partnership” between a disease advocacy group and a drug discovery company is novel. “The CF Foundation is taking an interesting step: This obviously has a high risk, but could also have a high payoff if it works.” By providing early support for the discovery of new drugs, Collins says, the foundation assures that the disease “will get more attention and more cutting-edge approaches than it would otherwise.” 4. BIOMECHANICS # Geckos Climb by the Hairs of Their Toes 1. Elizabeth Pennisi The Tokay gecko is the envy of every serious rock climber and Spiderman wannabe. This tropical lizard defies gravity, running up walls and upside down across ceilings as readily as across floors. It can hang from one toe pad—that's akin to holding oneself in midair by one fingertip. And that pad sticks to walls even in a vacuum and underwater. Gecko gecko's secret: rows of tiny hairs with multiple split ends on the bottom of each pad, says Kellar Autumn, a biomechanist at Lewis and Clark College in Portland, Oregon. While he was a postdoctoral fellow in Robert Full's lab at the University of California, Berkeley, Autumn figured out how these tiny hairs—each no taller and much more slender than the period at the end of this sentence—can be so strong. Armed with that knowledge, Autumn, Full, and their engineering colleagues hope to design synthetic “footpads” to improve the maneuverability of robots and perhaps to design an entirely new type of adhesive. As they report in the 8 June issue of Nature, weak attractive forces between the 1000 or so split ends on each hair and the ceiling help the gecko grab even the smoothest surface. Such forces are typically generated when two surfaces come very close together. Simply changing the angle of the hairs, called setae, causes these forces to disappear; then the ends let go and the lizard scrambles forward with no hesitation. “It's great to look at how evolution has solved mechanical problems,” marvels Bruce Jayne, a functional morphologist at the University of Cincinnati in Ohio. Gecko toe pads are covered with rows of setae made of keratin, the same protein in human hair and bird feathers. Each seta's curved shaft ends in many hundreds of spatulae, stubby tendrils—too small to see with a regular microscope—with rounded ends. To figure out how these hairs might help geckos hang upside down, Autumn tapped the expertise of Berkeley engineer Ronald Fearing and Stanford engineer Thomas Kenny. With the help of a microelectrical mechanical sensor, designed for use with atomic force microscopy, they were able to measure the lateral and perpendicular forces exerted by a single hair that had been removed from a gecko's foot. “The technical difficulty of measuring forces at such a small scale is really significant,” points out Jayne. That Autumn and his colleagues succeeded “is really admirable.” At first the hair didn't stick well to the sensor surface. But that changed after the researchers gently pressed the hair into the surface and then began to drag it across and nearly parallel to the sensor—movements that resemble how intact setae work as the gecko puts its foot down. With that motion, “adhesion is rapidly engaged, and that's when we see fairly large forces,” Autumn explains. Previously, Berkeley's Duncan Irschick had measured the overall adhesive forces of a gecko foot. From those, Full and Autumn had calculated the contributions of individual hairs, some 500,000 of which are arrayed in sets of four in a leaflike pattern on the pads. Amazingly, recalls Autumn, “each [hair] was 10 times more adhesive than we would have predicted.” One seta is strong enough to hold up an ant, and a million could support a small child. Autumn and other researchers have ruled out that suction, glue, or even electrostatic forces are responsible. Instead, he and Full think that as the spatulae get close enough to the surface, they generate weak intermolecular forces, akin to van der Waals forces, that sum to guarantee a secure foothold. “Geckos are way overbuilt,” explains Anthony Russell, a functional morphologist who has long studied geckos at the University of Calgary. That is how a gecko can cling to ceilings even though just a small fraction of its setae are oriented in an adhesive direction. Already Full's collaborators have built a robot gecko that scales walls and walks over obstacles. The current model uses pressure-sensitive adhesive and mimics how the gecko uncurls its toes as it puts its foot down and then peels the toes to detach the setae as it walks. These motions “reduce the attachment and detachment forces to almost nothing,” enabling the robot (and the gecko) to use up very little energy in the process, says Full. The next step is to outfit the robot with synthetic setae. The researchers don't expect to find a material that they can split 1000 times, however. Instead, they hope that studies of other lizards and also of kissing bugs, which have setae with few and sometimes only one spatula, will help them design simplified setae that can be manufactured. Eventually, Full and Autumn envision an all-purpose, reusable, gecko tape—one that leaves no residue behind. But Full is not so sure that gecko gloves and climbing shoes will ever be more than a rock climber's fantasy. 5. NEUROBIOLOGY # Trigger Found for Synapse Formation 1. Trisha Gura* 1. Trisha Gura is a science writer in Cleveland, Ohio. Because the ability to form connections between nerve cells is at the heart of all brain function, neurobiologists have looked long and hard for the molecules needed to achieve such biological hard-wiring. But they've had little luck in finding the matchmakers of nerve cell connections, called synapses, in the brain—until now, that is. In today's issue of Cell, molecular neurobiologist Tito Serafini and colleagues at the University of California, Berkeley, report that a single protein can apparently trigger synapse formation between brain neurons isolated from mice and grown in culture. The notion that just one molecule can jump-start so critical a process “is a big breakthrough,” says molecular and cellular neurobiologist Richard Scheller of Stanford University. If the finding is borne out in living animals, it could provide fresh insights into how the brain is wired during embryonic development and might eventually provide new ways to enhance or at least maintain synapse formation in the brains of patients suffering from neurodegenerative diseases such as Parkinson's or Alzheimer's. The critical players, according to Serafini's team, are either of two sister proteins called neuroligin 1 and -2. Neurobiologists had suspected that the neuroligins might play some role at the synapse since their discovery about 5 years ago by Thomas Südhof's team at the University of Texas Southwestern Medical Center in Dallas. At the time, the researchers were studying another set of proteins called neurexins that bind to a toxin from black widow spider venom. Because the neurexins can be produced in literally thousands of variants, investigators thought that the molecules might be involved in building the many different synaptic circuits of the brain—an idea buttressed by Südhof's finding that the proteins act like molecular glue to help cell surfaces adhere to each other. A neurexin, anchored in the surface of the transmitting cell of the synapse, would hook up to a binding partner on the cell that receives the connection, or so the thinking went. That hypothesis led Südhof to search for candidates that bind to neurexin. In 1995, his team pulled out neuroligin 1, and shortly after that, its relatives neuroligin 2 and -3. Using antibodies, the researchers also showed that neuroligin 1 is located in the synaptic membrane of the receiving neuron. But “the missing link,” says Südhof in retrospect, was showing that neuroligins can actually initiate synapse formation. Now, Serafini and his colleagues have done just that. The group, led by postdoc Daniel Emerling, first took mice and dissected out a set of brain neurons called pontine cells. The investigators then teased out a second group of neurons, granule cells, that connect with the pontine neurons in the brain. The researchers found that the two types of neurons form synapses in the petri dish, as they do in the intact brain. They could tell that the neurons were connecting, because they could see clusters of neurotransmitter-containing vesicles forming at the synapse. “We have this [granule] cell type that we know forms synapses very well with the pontine neurons,” Serafini says. “The beauty of our system is that we can mine it for the molecules involved in synaptogenesis.” Already the mine is yielding gold. Serafini, postdoc Peter Scheiffele, and their colleagues began by looking at a cadre of candidate genes for synapse formation that are active in granule cells. The researchers genetically engineered those genes, one at a time, into cells that would normally never steer synapse formation, such as human kidney cells or fibroblasts, and then they mixed the modified cells with pontine neurons in their culture system. After a series of genetic duds, the team hit pay dirt with the genes for neuroligin 1 and -2. Kidney cells that expressed either gene could trigger early synapse formation in the pontine neurons just as readily as their normal granule cell partners. At the outset, says Serafini, his team had hoped for nothing more than some slight changes in the pontine cells that would indicate they were beginning to form connections. “We never expected that a single protein, when expressed in multiple cell types, would drive the entire program of presynaptic development,” he says. Initially, many critics didn't believe the results either. Citing studies showing that plastic beads coated with a substance called polylysine also cause signs of early synapse formation, they argued that the neuroligins' effects might be nonspecific. But Serafini's group went on to address those concerns. For example, in what Serafini calls the “clinching experiment,” his team added a neurexin known to bind to neuroligin 1 to the cultured brain neurons. As hoped, they found a major drop in vesicle clustering—presumably because the neurexin bound to neuroligin 1 and hindered it from acting on the pontine cells. Synapse formation did occur, however, when they used a related neurexin that doesn't bind neuroligin 1. “These are very good controls for specificity,” says Scheller, who says he is convinced. “Now this molecule needs to be studied in more detail.” One way of doing that is to knock out the neuroligin genes in mice and see how that affects brain formation and function. Südhof's team already has a neuroligin 1 knockout. Although the rodents appear normal, Serafini notes that another neuroligin could easily pick up the slack for neuroligin 1. So the next step, he says, is to knock out all three neuroligin genes. Now that the Serafini team's assay has proved its mettle, the researchers plan to use it to look for more genes involved in synapse formation by brain neurons. “We are asking a big question, ‘What are the molecules that drive synapse formation in the central nervous system?’” says Serafini. As his team and others continue to “mine” nerve cells for their molecular precious metals, investigators are likely to gain a treasure chest of insights into how the brain develops and functions. 6. GERMAN SCIENCE # Max Planck Charts New Path 1. Robert Koenig Germany's premier basic research organization, the Max Planck Society, released a long-awaited blueprint for change during its annual meeting this week, recommending that the society's nearly 3000 scientists embrace more interdisciplinary and international projects in a range of new research priorities. In the half-century since Max Planck rose, reconstituted from the ashes of World War II, it has created a loosely knit empire of 78 institutes. Each institute is built around a handful of top researchers who have been given ample resources and considerable independence. Although that formula has produced excellent science—Max Planck scientists have won 10 Nobel Prizes since 1984—some critics contend that it has prevented the society from reacting quickly enough to sudden changes in the scientific landscape and has isolated its researchers from Germany's university system and from colleagues at other institutes (Science, 4 June 1999, p. 1595). View this table: Since becoming president of the Munich-based Society 4 years ago, biologist Hubert Markl has sought to address such concerns. He has, for instance, required more frequent outside evaluations of institutes, hired more researchers on short-term contracts, and developed “International Max Planck Research Schools,” which starting this fall will offer Ph.D. degrees in cooperation with German universities. The new blueprint, says Markl, will make the institutes “perhaps a little less independent, but much better interconnected with other research groups.” According to U.K. Engineering and Physical Sciences Research Chief Richard Brook, who led an international evaluation of Max Planck last year, the recent moves and the new report “indicate that Max Planck is moving in the right direction.” The report, called Max Planck 2000-Plus, is the product of an 18-month-long internal review. Its recommendations were formulated by some two dozen Max Planck researchers and administrators, who sought input from every institute. “We found a real spirit of innovation,” says Eduard Arzt, a director of the Max Planck Metal Research Institute who compiled the report's materials science section. Noting that “competitive, high-tech research is, in many cases, beyond the scope of one or several institutes,” the 2000-Plus report seeks to remedy that flaw by having scientists develop “an even more intensive collaboration” with universities and other research outfits. Toward that end, Markl says, Max Planck will soon launch the first nine International Research Schools for Ph.D. students, many from outside Germany. To cope with the data flood from “big science” efforts such as the international Human Genome Project, the report urges the rapid development of bioinformatics research at various institutes and multidisciplinary teams spanning several institutes and including other organizations. Some of this is happening already: Max Planck's astronomy researchers, for example, work closely with counterparts in Europe and North America to avoid costly redundant research. These changes, once implemented, could whittle away a researcher's ivory tower independence. But Markl thinks the trade-off—more collaborations and an influx of young minds—will spur a new era of creativity. “The most important driving force,” he says, “will be the increasing mobility of scientists, especially young researchers, across borders and among institutions.” 7. PALEONTOLOGY # New Feathered Dino Firms Up Bird Links 1. Dennis Normile BeijingVolcanic eruptions some 125 million years ago entombed a menagerie of ancient animals at a site in northeastern China that is proving to be a treasure trove for paleontologists. It has also become ground zero for the continuing debate on the origins of birds. Last week, Chinese scientists presented evidence from a new specimen dug up in Liaoning Province (Science, 13 March 1998, p. 1626) that they say strengthens the case for a link to dinosaurs—and for the value of further work at the site. The finding, one of several fossils displayed at a meeting here,* is the third known specimen of a strange creature known as Caudipteryx. When the first Caudipteryx was discovered in Liaoning in 1998, most paleontologists classified it as a member of a group of two-legged, carnivorous dinosaurs known as theropods (Science, 25 June 1999, p. 2137). Unlike any other known dinosaur fossil, though, the tail and stubby forelimbs of Caudipteryx showed the unmistakable imprints of feathers—features most paleontologists believed it had inherited from a dinosaurian common ancestor it shared with birds. However, Caudipteryx appears to have been earthbound, lacking a full set of wing feathers and other features of wings. The new specimen was described by Zhou Zonghe, a paleontologist at the Chinese Academy of Sciences' Institute of Vertebrate Paleontology and Paleoanthropology (IVPP) in Beijing. Although the fossils reveal some beautifully preserved feather impressions, “it's the bones that are important,” says Zhou. The fossil lacks a head, but the rest of its skeleton (see picture) is better preserved and better articulated than those of its two predecessors. Zhou admits that the bones show a number of birdlike characteristics usually absent in dinosaurs, including a thumblike appendage for perching. But he has identified 16 characteristics more similar to dinosaurs than to early birds, including the proportions of the bones in the foot and the shape and orientation of the pelvis and bones in the pelvic region. He has labeled the new species Caudipteryx dongi, after prominent Chinese dinosaur expert Dong Zhiming. Zhou's work reinforces the views of most Western scientists. “The Chinese finds are [illustrating] very nicely the transition between things that are true dinosaurs to fully avianlike creatures,” says Luis Chiappe, a paleontologist at the Natural History Museum of Los Angeles County. University of Chicago paleontologist Paul Sereno agrees that identifying more dinosaurlike features “is big news” that lends support for the idea that there is no sharp evolutionary dividing line between dinosaurs and birds. A minority remain unconvinced, however. The feathers on Caudipteryx are considered less evolved than those on Archaeopteryx, the oldest commonly acknowledged bird, notes Storrs Olson, an avian paleontologist at the Smithsonian Institute in Washington, D.C., yet the Liaoning fossils are 25 million years younger than Archaeopteryx. And although Caudipteryx may have had true birdlike feathers, he says, that doesn't make the case for other recent Liaoning finds such as Sinosauropteryx, whose hairline “protofeathers” he dismisses as “fuzz.” The Liaoning sites, he adds, “are showing us kinds of birds we didn't know a whole lot about. But there is no information about the origin of birds.” Zhou disagrees, saying that placing Caudipteryx with the birds would require its bird ancestors to have evolved an implausible number of dinosaurlike characteristics. It is more likely, he says, that the creature's handful of bird characteristics are either due to parallel evolution or appeared on the way to the origin of birds. The dinosaur-to-bird believers think it can only be a matter of time before more of the missing links are uncovered. “We are convinced that birds evolved from theropod dinosaurs,” says Ji Qiang, director of China's National Geological Museum in Beijing. “But we don't know yet from just which group [of theropods].” He and others hope that the ground will yield more definitive answers. And they agree that the most logical place to dig is in Liaoning Province. • *5th International Meeting of the Society of Avian Paleontology and Evolution and the Symposium on Jehol Biota, IVPP, Beijing, China, 1 to 4 June. 8. ANIMAL CLONING # Clones: A Hard Act to Follow 1. Elizabeth Pennisi, 2. Gretchen Vogel Despite scores of cloned animals, the process is fraught with problems. Many researchers are going back to the lab to find out why Cumulina. Cupid. Peter. Webster. Diana. Dotcom. Dolly. Once the realm of science fiction, cloned animals are now becoming almost commonplace. In the past 4 years, cows, mice, goats, and pigs have joined sheep in an expanding menagerie of cloned mammals. Just around the corner, to judge from the press releases and headlines, looms the next brave new world of biotechnology, with herds of identical cattle, sheep, and goats producing bucketfuls of drugs in their milk, pigs designed to grow spare parts for humans, primates and other animals custom-cloned to study human diseases, and even replacement parts cloned from a patient's own cells. Indeed, given the seemingly endless string of birth announcements, a logical question is, “Will humans be next?” Not likely, say numerous experts in the field. What the press accounts often fail to convey is that behind every success lie hundreds of failures—some so daunting that many would-be cloners have put efforts to create live animals on hold and are going back to the lab to study why cloning sometimes works but far more often fails. Despite years of effort, “we're in the same bind that we've always been in. A majority of [would-be clones] do not make it to term,” says Robert Wall of the U.S. Department of Agriculture (USDA) in Beltsville, Maryland. “We have no explanation; it's more art than science,” adds Jean-Paul Renard of the National Institute of Agricultural Research (INRA) in Jouy en Josas, France. Indeed, even Ian Wilmut, the Scottish researcher who brought the world Dolly, hasn't cloned another animal in years; instead, he is trying to find out what makes cloning by nuclear transfer possible by studying how genes are reprogrammed. Although Wilmut, who works out of the Roslin Institute near Edinburgh, isn't throwing in the towel, he says enormous hurdles must be overcome before cloning becomes practical, much less profitable. First and foremost is the problem of efficiency, which remains at a less-than-impressive 2%; out of some 100 attempts to clone an animal, typically just two or three live offspring result. Even when an embryo does successfully implant in the womb, pregnancies often end in miscarriage. A significant fraction of the animals that are born die shortly after birth. And some of those that survive have serious developmental abnormalities, suggesting that something in the recipe is fundamentally wrong. What's more, cloning, however arduous, is just the first step. If the goal is to create “bioreactors” that produce therapeutic proteins in milk, or pig pancreases the human body will not reject, then cloners need to insert foreign genes into the genome in exactly the right place—a process that has so far defied most efforts. “The issues are back in the laboratory rather than in the barnyard waiting for something to gestate,” says Wall. Cloning veteran Jim Robl of the University of Massachusetts (UMass), Amherst, agrees. Since he and his colleagues at the biotech company Advanced Cell Technology (ACT) in Worcester, Massachusetts, cloned six transgenic calves, they have focused less on producing live offspring than on questions such as whether cloned animals are genetically older or younger than normal (Science, 28 April, pp. 586 and 665). To increase their odds of producing healthy clones, researchers are now probing fundamental questions of cell biology, such as what kinds of cells make the best donors or what environments are most conducive to the earliest stages of development. They are trying to figure out whether there is something inherently flawed in “asexual” reproduction in mammals—in other words, do we really need two parents? Researchers know that genetic competition between sperm and egg helps to modulate imprinting, a process that selectively silences certain genes early in development. That process may go awry in cloning, accounting for some of the developmental abnormalities. Or does some problem lie in the “in vitro” component? Toward that end, teams are sorting out more mundane questions of how, exactly, to culture the growing embryo in the lab, and what concoction of hormones is necessary to ensure adequate development. In all species, the basic hurdles are the same, but the details differ sufficiently that each species has gotten sidetracked at different points along the way to becoming a commercially or medically useful clone (see chart on facing page). Yet in the highly competitive world of animal cloning, researchers are loath—or sometimes forbidden—to share their tricks of the trade. Added to the normal passions, jealousies, and simple desires for credit that plague most high-profile research is the fact that much cloning research is done with corporate sponsorship—and with corporate requirements of secrecy. “Breakthroughs” are often announced long before the technical details are published in journals, making it hard for researchers to verify or extend the results. Even attracting scientists to a recent closed-door conference* at the Banbury Center at Cold Spring Harbor Laboratory in New York required a “major diplomatic effort,” says molecular biologist Norton Zinder of The Rockefeller University in New York City, who helped organize the meeting to try to promote better communication among the cloners. Dolly and friends With new clones being announced almost monthly, it is easy to forget just how mind-boggling the process really is. Take a single adult cell, whose fate is supposedly sealed, and send it back in time, so to speak, unsealing the genetic instructions contained in that cell's nucleus. Then ask that nucleus, once inserted into another cell, to set that cell on a course of replication and differentiation to produce a whole new animal—one that is a veritable carbon copy of the adult from which that cell came. There's no true biological “mother” or “father” involved. Through the past half-century, researchers had dabbled with this bold endeavor, transferring nuclei from a variety of cell types and sources (see this month's “Pathways of Discovery” essay on p. 1775) into cells whose own genetic material had been removed. Sometimes the nuclear transfer experiments seemed to work. In cows, for example, when nuclei from relatively undifferentiated embryonic cells were put into ripened eggs ready for fertilization, offspring could result. But cloning from an adult or even a fetal cell, which would have begun to differentiate, seemed impossible. To pull off their paradigm-altering experiment, Wilmut and Keith Campbell at the Roslin Institute spent years painstakingly manipulating both the donor cells and the receiving eggs, developing the finesse to make nuclear transfer work with differentiated cells. Ultimately, they teamed up with PPL Therapeutics of Midlothian, Scotland, which wanted to make herds of identical sheep that carried a human gene for a therapeutic protein. The Roslin group thought they might succeed where others had failed if they could synchronize the cell-division cycle of donor cells with that of the egg. They did this by depriving the donor cells of nutrients, which caused them to shut down most genetic activity. At the same time, they worked out better ways of removing the DNA from the ripened oocytes and, after fusing the oocyte with the donor cell, of triggering cell division as if the egg had been fertilized. Wilmut, Campbell, and their Roslin colleagues first tried their idea out with embryonic cells, which were presumed to be more malleable. Their success in cloning two lambs in 1996 prompted them to try the same approach using older cells—eventually, cells that they had cultured from an adult ewe's udder. That the experiment worked with the mammary cells was just short of miraculous: Dolly was the product of 434 attempts at nuclear transfer, all but one of which went bad. But that one lamb, reported in February 1997, was enough to set off a worldwide tizzy. Not only did Wilmut's work demonstrate, for the first time, that specialized cells might be reprogrammed and revert to the time when they could become any and all cell types, but it also implied that, if the process worked in sheep, then humans might be just around the corner. Several colleagues were skeptical, suggesting that perhaps Dolly was the product of a rogue fetal or undifferentiated stem cell and not a true mammary gland cell. But others were in awe and rushed off to try to replicate the results in a variety of animals, from mice to cows. Within a year, companies like ACT and the newly formed Infigen in DeForest, Wisconsin, made public their own successes with cattle, and cloning became a household word. Meanwhile, Campbell moved over to PPL and set out to take the next step: putting foreign genes into the donor cell's DNA. It didn't take much work to add new DNA to the cultured fetal cells, select those that took in the new genes, and fuse them with enucleated eggs. It worked, as evidenced by the birth of Polly and five other lambs bearing the gene for human factor IX, published in December 1997. Yet even though Polly expressed the new gene—that is, made the protein encoded by the gene—she was just an interim success. Campbell had inserted the gene into a random location in the donor cell's genome, which meant he had little control over how active it would be. To guarantee high production of factor IX protein in sheep's milk—essential if the research was ever to yield a commercially viable bioreactor—Campbell needed to target the gene to a specific spot in the genome. But there was a problem. Until then, gene targeting had only worked in mice and, in one experiment, in human connective tissue cells—and decidedly not in livestock. That would soon change. In 1997 David Ayares, a molecular biologist at a pharmaceutical company, joined PPL with one goal in mind: gene targeting. The big problem, he, PPL's Alex Kind, and their colleagues quickly surmised, was finding a way to insert the gene before the donor cells got too old. Most cells can divide only a limited number of times in culture, and gene targeting requires several steps in which DNA is inserted and the few select cells that incorporate it into their chromosomes are allowed to multiply until there are enough for the next modification. “By the time you go through and get a large enough population, it's really pushing the limits [of the cells' ability to divide],” says Robl of UMass. PPL researchers on both sides of the Atlantic set out to improve the efficiency of each step. They also concocted different brews of growth factors that kept the cells healthier through more cell divisions. In August 1999, at a meeting on transgenic animals, Ayares showed the fruits of this labor: slides of Cupid and Diana, two sheep clones, one containing a marker gene as a control, and one containing both the marker and a gene for alpha-1 antitrypsin, a potentially therapeutic human protein. Both genes were “knocked” into the sheep's genome in much the way gene targeting is done in mice—in other words, they were precisely inserted into the correct spot. Ayares's announcement was “the most earthshaking news of the year,” recalls Wall of USDA, as gene targeting in livestock no longer seemed to be an insurmountable problem. The work has not yet been published, however (it is scheduled for publication in Nature), and no one has yet repeated the results. Even so, Robl is confident that others will soon succeed. As for PPL, Ayares says the team has pulled off the same feat in cow and pig cells—and that it's just a matter of time before PPL will turn those cells into clones. Cows: 200 and counting Wilmut's success in part grew out of the failures of nuclear transfer in cattle. Wilmut and Campbell built on a technique that researchers had been using for years to “clone” cows from very early embryonic cells. Indeed, in the 1980s, a Texas ag-biotech start-up named Granada had built its business plan on cloning fast-growing cattle this way. At the time, recalls Ken Bondioli, now with Alexion Pharmaceuticals in New Haven, Connecticut, nuclear transfer in cattle had become routine. Although the efficiency was low, “we produced hundreds of calves,” he says. But there was a catch. More than the usual number of pregnancies were proving problematic: Deliveries were difficult, and many calves died just before or after birth. “It took us some time to recognize this as a problem having to do with nuclear transfer,” Bondioli explains. Eventually, their data led Granada scientists and others to characterize what they called “the large calf syndrome,” the cause of which remains a mystery. View this table: Unable to overcome the problem, the company shut down by 1991. Not until Wilmut and the Roslin group produced Dolly and ACT and Infigen had cloned calves with genes randomly inserted into their genomes did nuclear transfer in cattle seem attractive again, says Bondioli. Very quickly, cloning successes in Japan and New Zealand showed that nuclear transfer of fetal or even adult cells worked just fine in cows—or so it first appeared. But as the worldwide total of cloned cattle approaches 300, researchers are finding that they, too, are haunted by Granada's ghost: About a quarter and sometimes more of the calves that survive to birth are bigger than normal, and many of the frustrating spontaneous abortions involve fetuses that are unusually large. Even the normal-sized newborns frequently have lungs like those of premature babies. Others seem to have blood potassium levels high enough that “the calf should be dead,” says Michael Bishop of Infigen. Adds Randy Prather of the University of Missouri, Columbia, “Just because you've got offspring doesn't mean they are normal.” In trying to sort out what goes awry, researchers are focusing on two areas. One is imprinting, the critical but poorly understood process by which the protein signal is determined by whether a certain gene came from the mother or the father (Science, 25 September 1998, p. 1984). Developmental abnormalities result when one copy of the imprinted gene turns on or shuts down inappropriately—a likely prospect in clones, as the whole genome comes from one donor cell rather than the typical two. Another possibility is that problems arise because of the way the egg is handled before implantation—for instance, if the brew of hormones is not quite right, or if the jostling, poking, and prodding damages the egg in some imperceptible ways. One hint is that cattle conceived in test tubes tend to have some of these same abnormalities. Robl of UMass wonders if they are stymied by both problems: mishandling of the egg and a lack of reprogramming of the donor nucleus. “What we still have is a black box,” he admits. He and others are now systematically evaluating each step. For instance, Bishop and his colleagues at Infigen are keeping a comprehensive database of cell lines, the specific techniques used during nuclear transfer, care and feeding of the surrogate mother, and in utero growth rates in an effort to try to predict which calves will have higher risks of problems. Already they have noticed a correlation between a certain cell line and unusual fluid buildup around the placenta. The company is also using microarrays to assess which genes are active in a given cell to see if they can discover a connection between certain genes and cloning success. “In 10 years, I'm sure we'll look back and see how archaic we are,” Bishop predicts. The not-so-impossible pig This March at the Banbury Center meeting, Prather likely experienced one of the worst moments in his career. After almost three frustrating years of trying to clone pigs, he was close to calling that goal unachievable. Hiroshi Nagashima of Meiji University in Tokyo, another speaker that afternoon, had a similar tale of woe. But just before their session was to begin, Alan Colman of PPL Therapeutics made a startling announcement: “I hate to have to say this, given what's coming up, but we've got pigs.” A press release about those five piglets, which included a randomly inserted transgene, made the headlines a few days later, although a peer-reviewed scientific paper has yet to be published—convincing most would-be pig cloners that their long-sought goal is now in reach. Pigs are one of the hottest commodities in cloning, as many scientists believe they are the key to xenotransplantation. Because of their size, pig organs are considered most likely to be compatible with humans and could thus satisfy the unmet need for replacement organs, such as hearts or pancreases. Moreover, some researchers think pig tissue, transplanted, say, in the brain, might be a source for much-needed chemicals that the human body fails to make, such as dopamine, whose loss contributes to Parkinson's disease. So it was no surprise that several animal scientists redirected their research toward cloning pigs once Dolly burst on the scene. The problem is that until 1998, few scientists had tried to work with immature pig eggs or to grow pig embryos in the lab. Pigs differ from cows and sheep in that they are born in litters, and unless there are at least four viable fetuses in the womb, the pregnancy fails. That means that a day's work has to yield at least several viable embryos if the cloning experiment is to have any chance of success. Ayares says PPL spent more than a year trying to clone pigs with the techniques the company and Wilmut had used for sheep. Each attempt failed. Pig embryos proved too fragile, and the cells often broke apart during nuclear transfer or handling. In those rare instances when the researchers were able to add the donor nucleus to the egg and then activate development, the embryos never made it to the blastocyst stage. Then 2 years ago, company scientists junked that approach and hit upon an entirely new—and apparently successful—one that Ayares will not discuss until the scientific paper comes out—other than to say that he is confident that PPL can clone more pigs when they want to. Next time around, he says, the company plans to genetically modify the donor cells to make pig organs more acceptable to the human immune system—a key step toward making xenotransplantation a reality. Prather, however, is withholding judgment until he sees more piglets. “I think they got lucky,” he says of PPL. Filling out the barnyard Cloners are testing the waters on other livestock, with mixed success. Goats, it seems, are easy. In 1999, two companies reported—one through a press release and the other in Nature Biotechnology—that they had successfully cloned goats. And this year, Nexia Biotechnologies Inc. of Montreal announced the birth of two goats, Webster and Peter, that carry a gene from arachnids that codes for the spider silk protein. This spring, says a company spokesperson, Nexia mated the two bucks with normal females; by year's end they expect the female offspring will be churning out milk chock-full of spider silk protein. The company plans to extract the protein and spin it into light, high-strength fibers for use as sutures or in bulletproof vests or automotive and aerospace components. On the other hand, despite their natural fecundity, rabbits have so far defied efforts to produce them in the lab. “We can get a lot of cloned embryos,” says Renard of the National Institute of Agricultural Research in France, but all pregnancies abort after transfer to a surrogate mother. He suspects that the problem may be in the earliest cell divisions in the embryo. Renard and his colleagues will keep trying, however, as transgenic rabbits produced by cloning could be valuable tools for studying cardiovascular disease. In terms of bioreactors, it would be tough to beat the chicken—or more precisely, its egg, says USDA's Wall—which is why several companies are now trying to clone chickens. A typical egg costs about 2 cents to make, and if cloners can insert a therapeutic gene and get it to express in the egg white, commercial technology already exists for separating the whites from the yolks. But the eggs themselves present cloners with a distinct problem: their huge size, says Leandro Christmann, a reproductive biologist at AviGenics in Athens, Georgia. And size does matter. To remove DNA from mammalian eggs with diameters of roughly 100 micrometers, you simply put the fairly transparent egg under a microscope and suck out the DNA with a pipette. But in chickens, the egg yolk is far too big and opaque. Just figuring out how to “see” the chicken egg nucleus has been a challenge, Christmann says. Even so, the start-up says it has made progress in developing or adapting the technology to take that egg through all the necessary steps, prompting AviGenics president Carl Marhaver to predict that within a year, “we will be able to produce the first cloned bird.” If primates, then humans? Perhaps no area of cloning research evokes more curiosity than primates. Although researchers aren't attempting primates as a dry run for humans—their goal is to create identical animals to study such diseases as hepatitis—their progress is likely to shed light on when it might be technically possible to clone people. Those worried that some crazed scientist might ignore the ethical and legal sanctions against human cloning experiments and plow right ahead can rest assured: It won't be easy, at least according to Tanja Dominko, a reproductive physiologist at Oregon Health Sciences University in Portland. When she arrived in Oregon in 1997, prospects looked fairly bright. Oregon's Don Wolf had just succeeded in using nuclear transfer to produce two monkeys, Neti (Nuclear Embryo Transfer Individual) and Ditto, a year earlier. As Granada had done with cows in the 1980s, Wolf had used nuclei from embryonic cells—far easier to work with than adult cells. But 300 attempts and no pregnancies later, the picture “is not as rosy,” Dominko says. Cloning primates “is not just around the corner.” Neither she, working with Oregon's Gerald Schatten, nor Wolf's team working one floor below, has been able to replicate Wolf's early success. Once Dominko realized what she was up against, she tried to determine whether the problem was with nuclear transfer or the in vitro procedures. She attempted “mock” nuclear transfers, in which she and her colleagues went through the cloning procedure but didn't actually replace the egg's own DNA. Instead, they fertilized the egg in vitro after poking and prodding it the way they would have for a true nuclear transfer experiment, and then placed it in a female for gestation. Those attempts didn't work well, suggesting to Dominko that the in vitro procedures were the problem. She then made “egg-friendly” improvements such as using sperm extract instead of harsher chemicals to prompt the egg to divide, which helped the subsequent nuclear transfer experiments. The team produced roughly 45 embryos by nuclear transfer this way, but none successfully implanted in a surrogate female monkey's womb. Then Dominko, Schatten, and colleagues began looking at the transferred nucleus itself. Under a light microscope, the embryo's expanding cluster of cells, known as the blastocyst, looked just like those seen after successful in vitro fertilization. But a closer look at these dividing cells, with confocal microscopy, revealed “a whole gallery of horrors,” says Schatten. The new nucleus seemed completely out of sync with the egg. Even the first cell division had gone awry, as the chromosomes didn't seem to have copied and separated as they should have. By the eight-cell stage, some cells had too much DNA, while a few seemed to have none at all. Dominko and Schatten then took a closer look at the spindle, and in particular at the centrosomes, which help organize and guide the movement of DNA during cell division, making sure that each cell gets the right complement of chromosomes. In primates, they found, the incoming nucleus tends to leave behind one or both of its centrosomes. “The embryos we were making probably never had a chance,” says Dominko. Given these results, which are still unpublished, Schatten and Dominko have all but given up on cloning by nuclear transfer until they develop a better understanding of these abnormalities. Instead, they have turned to embryo splitting, in which the early embryo is divided in two and gives rise to identical twins, as a means of generating like animals useful for research (Science, 14 January, p. 317). Dominko and Schatten don't know why primates are different from cows, but they are convinced that attempts to clone humans would run up against the same biological roadblocks. Even Wolf on the floor below is now looking at embryo splitting, but he has not abandoned nuclear transfer. He's tried nuclear transfer with some 100 embryos, none of which has established a pregnancy. Indeed, his studies have revealed another source of failures: Embryos don't develop at the same rate in a lab dish as they do in the womb. It's important to keep trying, he argues, as clinical studies often require more than the two identical animals that can be produced by embryo splitting. Litters and litters of mice Despite efforts by numerous labs to clone mice, this laboratory staple has proved remarkably elusive. Indeed, until recently, only one person in the world had been able to pull it off: Teruhiko Wakayama, who originally reported success with Ryuzo Yanagimachi at the University of Hawaii, Honolulu, in July 1998. Even in Wakayama's skilled—and some say “magic”—hands, cloning is unpredictable and enigmatic. Fatal problems can crop up at every step. Even the temperature of the lab can make a difference: Slightly too hot or cold, and the technique won't work as well. When Wakayama first moved from Honolulu to Rockefeller University in November 1999, nothing seemed to go right. For the first few months, few of the embryos survived, says neuroscientist Peter Mombaerts, who helped lure Wakayama and his colleague Tony Perry to Rockefeller. One contributing factor, the team suspects, was that their brand-new incubator was producing toxic gases and killing the embryos. Even without toxic incubators, other labs remained frustrated in their efforts to duplicate Wakayama's work, prompting some disbelief. Wakayama and Yanagimachi's procedure involves injecting the nucleus into the enucleated oocyte instead of fusing the entire donor cell with an electrical charge, and this feat requires the steady hand of a surgeon. “It requires very miniature handwork, and it has to go reasonably fast,” explains Mombaerts. Adds Perry: Wakayama “certainly has the magic touch.” But now others working with Yanagimachi and at least three new labs are claiming to have the magic touch as well. Atsuo Ogura of the National Institute of Infectious Diseases in Tokyo and his colleagues have cloned mice with the Honolulu technique, and a description of their work was published this month in Biology of Reproduction. Renard says his group in France has also produced a few litters, with several more pregnancies under way. And after tutoring from the Hawaii team, postdoc William Rideout and graduate student Kevin Eggan in Rudolf Jaenisch's laboratory at the Massachusetts Institute of Technology have also successfully repeated the technique. Most agree that Wakayama has a knack for the injections, but “the technique is transferable,” Jaenisch says. That is good news for the cloning field as a whole, Jaenisch says. Scientists are eager to use cloned mice as a powerful lab tool—not the least to study cloning itself. Because mice are small and reproduce quickly, and because scientists know so much about their genetics and their development, researchers say the mouse offers the best hope for answering many of the questions that plague efforts to clone other species. One of those other species, of course, is humans. For now, the serious obstacles to cloning every species, especially other primates, suggest that human cloning—even so-called therapeutic cloning to produce cell lines to be used in treating disease—may be a long way off. As for reproductive cloning, or actually creating a living replica, “it would be criminal at this stage in our abilities,” says Zinder of Rockefeller. Most researchers concur. The U.S. National Bioethics Advisory Commission issued a report in 1997 saying that human reproductive cloning would be unethical for a variety of reasons, and the commission's chair, Harold Shapiro of Princeton University, says it is still “clinically and scientifically premature to produce human infants.” Even if the technique were safe, he adds, it would be unethical to proceed without a clearer public consensus. Given the huge scientific unknowns, there should be ample time for sorting out whether human cloning would ever be acceptable should it, too, yield to the magician's touch. • *Mammalian Cloning, Biology and Practice was held 12 to 15 March in Cold Spring Harbor, New York. 9. ANIMAL CLONING # Profits From Precious Pets 1. Elizabeth Pennisi* 1. With reporting by Dennis Normile. Face it, dog owners are suckers when it comes to their pets. Likewise, cat enthusiasts. So it should be no surprise that well before the first dog or cat is cloned, scientists-cum-entrepreneurs are already cashing in on this unconditional affection. Some four companies have set up shop storing tissue from people's favorite pets until the time is ripe to clone them. Each offers advice and a retrieval kit that veterinarians can use to collect skin, mouth, blood, or mammary cells from a living or even recently deceased animal. The companies culture the tissue until there are several million potential donor cells for nuclear transfer. Then the vials of cells are slowly frozen and stored in liquid nitrogen freezers until science advances to the point when Fido can be reproduced from scratch. The boom in frozen-tissue storage started in 1998, after an anonymous millionaire, hoping to clone his pet dog Missy, awarded Texas A&M University animal scientist Mark Westhusin$2.3 million to develop the necessary techniques. Shortly thereafter, PerPETuate Inc. started up in Farmington, Connecticut, and about the same time, Canine Cryobank, a San Marco, California, company already in the business of transporting frozen semen of high-priced pets and show dogs, began freezing tissue for future cloning as a sideline. Richard Denniston, an animal scientist at Louisiana State University in Baton Rouge, started Lazaron BioTechnologies there, in part because his department was getting so many calls from pet owners requesting this service, he says. The last one on the market, Genetic Savings and Clone, is connected with the Texas A&M group working on cloning Missy in the much-publicized Missyplicity Project. Its doors opened for business earlier this year, in part because “these other companies started to cash in on our investment,” says Westhusin.

Each company charges between $300 and$2000 for the tissue-retrieval kits and cell preparation procedures and then tacks on storage fees adding up to about $100 per year. Storage fees are likely to add up: Right now, even the most optimistic cloning enthusiasts think dog and horse cloning is still several years away. Prospects for cats look somewhat brighter, although significant challenges remain. Consider Missy. Three years into the Missyplicity Project, Westhusin knows all too well how tough his task is. The reproductive biology of dogs makes them more difficult to manipulate than any of the animals cloned to date. “They go 6 months between ovulation cycles, so you have to have a huge kennel of animals if you expect to [try cloning] on a daily basis,” says Jim Robl, a reproductive biologist at the University of Massachusetts, Amherst. Thus, even though Westhusin is now working with 60 dogs—some as donors, others as surrogate mothers—when he does have an egg ready for nuclear transfer, the chances are that he won't have any females primed to receive an embryo, should he succeed in making one. Just getting started has been a challenge. Westhusin first tried the techniques he used with cows to prompt immature dog eggs to mature, but to no avail. At this point, he collects lots of immature eggs from spaying clinics, but only every once in a while does he get one to mature enough to be able to start dividing, assuming it had accepted the donor DNA. And when the researchers get a mature egg, there is no time to spare. No procedures exist for keeping the egg healthy once it has taken on its new DNA and begun to divide, so the Texas team must put it into a receptive female within hours of the nuclear transfer. Needless to say, “we have no clone pregnancies,” Westhusin told Science in April. Cats, however, are another story. Even though no millionaire has tried to jump-start cloning his favorite kitty, cats are much more prolific, with a 2-month gestation time and frequent ovulations during a year. Would-be cat cloners can also draw on the wealth of knowledge about feline reproductive physiology, garnered mostly from research on endangered wild cats, says Cornell University's Jonathan Hill. As a result, researchers know what hormones will make female cats come into heat and also how to culture feline embryos outside the womb. In March, Tatsuyuki Suzuki of Yamaguchi University in Japan set cat cloners' hearts aflutter when he announced he had used nuclear transfer to clone an embryo from a skin cell of a dead cat. At the time, he expected he would have a live cat clone by June. The Japanese team has since tried twice to implant eggs, with no success. Several groups in the United States have forged ahead, and at least one, Advanced Cell Technology (ACT) in Worcester, Massachusetts, has had no trouble establishing pregnancies, says ACT vice president of scientific development Robert Lanza. Of course, that's no guarantee he will soon have a healthy kitten clone, as establishing pregnancies is only part of the problem (see main text). Nevertheless, “we expect to be first,” asserts Lanza, although even he is hesitant to predict when. Whichever pet is cloned first, there is of course no guarantee that the animal will have the same loveable personality as the donor. And the cost is likely to be prohibitive, as high as$200,000 per animal. Eventually, the companies predict, costs should drop to a more modest $20,000, or perhaps even$5000 a pop. That might be a good deal for biomedical researchers seeking to clone dogs for use in studying human diseases. But that cost is still a good deal more than a trip to the pound.

10. PALEONTOLOGY

# Learning to Dissect Dinosaurs--Digitally

Technology borrowed from medicine and industry helps paleontologists peer into specimens with x-ray eyes

Chris Brochu's mission was to get inside the head of the most expensive dinosaur on Earth. Brochu, a paleontologist at The Field Museum in Chicago, was hired to describe the complete anatomy of Sue, an $8.4 million Tyrannosaurus rex—including the hidden cavities of the brain. As with any precious specimen, cracking open the skull was clearly out of the question. Instead, Brochu and his colleagues shipped the 1-ton skull to Rocketdyne, a division of Boeing in Chatsworth, California, which put Sue in a computed tomography (CT) scanner designed to find microscopic flaws in rocket engines. As the skull slowly rotated, x-rays penetrated and created salami-like cross sections. Back at The Field Museum, Brochu digitally manipulated the scanned images to produce a virtual cast of Sue's braincase that revealed olfactory bulbs the size of grapefruits. The result, Brochu reported in the March issue of the Journal of Vertebrate Paleontology, bolsters the idea that the enormous predator had a well- developed sense of smell. Physicians routinely use CT scanning to diagnose brain tumors and other life-threatening conditions. But in the past 2 decades, as the images have grown sharper, paleontologists have increasingly trained the technique on long-dead animals as well. CT cross sections offer an inside view of rich anatomical detail, including everything from bony canals that once held nerves to the internal networks of dinosaur noses. Off-the-shelf software allows any scientist to view the slices, while more sophisticated and expensive programs generate a three-dimensional (3D) replica of a fossil. The resulting virtual specimen can help preparators chisel a fossil from its matrix without destroying any bone and can reveal features that even the most delicate preparation could never uncover. And digital data, unlike priceless specimens, can be easily and cheaply transported to colleagues. CT scanning is “a very important tool for paleontology,” says Jack Horner, a dinosaur expert at the Museum of the Rockies in Bozeman, Montana. To be sure, the technique won't work on all samples. And the images still require a trained eye to interpret. Still, CT scanning has contributed to a number of high-profile discoveries, the latest of which may be a four-chambered dinosaur heart (Science, 21 April, p. 416). For some paleontologists, using the tool can be as much fun as being in the field. “It's a very exciting feeling, like when you're excavating a fossil,” says Larry Witmer of Ohio University College of Osteopathic Medicine in Athens. “When you look inside the brain cavity of a dinosaur skull encased in rock, you are the first person to look inside its head.” Inside view One of the first scientists to experience that thrill was Glenn Conroy, an anthropologist who was then at Brown University. In 1983, while waiting in a supermarket check-out line, he saw a magazine cover with a computer-generated image of a human face. Craniofacial surgeons were using the then-novel technique—producing 3D images from CT scan digital data—to help reconstruct the face and skull of a 4-month-old girl with a birth defect. “I was amazed that you could electronically dissect a living child's face,” Conroy recalls. “It gave me the idea that maybe we could do this with fossils.” Conroy teamed up with the doctors who had scanned the child's head, a group at Washington University School of Medicine in St. Louis, where Conroy now works. They tried out the CT scanner on a common fossil ungulate called an oreodont. The cranium was filled with sandstone, which Conroy hoped the CT scan could distinguish from the fossil bone. Standard medical x-rays aren't always sensitive enough to show the contrast between bone and rock, and they make 3D objects such as skulls difficult to interpret. “When you put something in an x-ray machine, you're looking at everything at once,” Horner says. “It's hard to tell what's on top, what's on the bottom, and what's in the middle.” In computed tomography, however, the x-ray source rotates around a specimen, probing one cross-sectional region at a time with a fan-shaped beam. Sensors measure the beam's attenuation, which varies with density. The result, Conroy found, was a 3D image that neatly distinguished between fossilized bone and the rock matrix. “It worked beautifully, given the primitive state of scanners,” he says. Even so, several stumbling blocks remained. Medical scans were designed to be viewed on a monitor or film, not downloaded to a computer. Moreover, early scans were grainy—about on par with early Pac-Man video games. “It was almost a joke,” says Jim Clark, a paleontologist at George Washington University in Washington, D.C. Eventually, the images became good enough for clinicians, but still not sharp or powerful enough for paleontologists to study small, finely detailed fossils. Paleontologists turned to industrial scanners, which are designed to probe important mechanical parts for flaws. These machines yielded better resolution and thinner slices; so-called micro-CT scanners can now obtain slices down to 10 micrometers. In 1993, Tim Rowe, a paleontologist at the University of Texas, Austin, showcased the technology on a rare and fragile 3.8-centimeter-long skull of Thrinaxodon liorhinus, a 240-million-year-old relative of early mammals. He had struck up a collaboration with Scientific Measurement Systems, an Austin-based company that builds scanners with a resolution two orders of magnitude higher than that of most medical scanners. “The first slices that came out made my eyes bug out,” Rowe recalls. The result was a CD-ROM that many paleontologists consider a landmark digital monograph. Another advantage of industrial scanners is that they can handle fossils much bigger than the human body. In 1992, Horner and his team loaded up all of his museum's dinosaur skulls—including ones that wouldn't fit into a medical scanner—and drove them to Cincinnati in a U-Haul truck. There they CT-scanned the fossils at a facility where the General Electric company examines jet engines. Horner now has two UNIX computers crammed with CT scans from some three dozen dinosaurs. Paleontologists such as Horner scrutinize CT images for traces of long-vanished physiology. Ohio's Witmer, for example, found extensive networks of capillaries in the nasal cavity of several dinosaurs that may have helped cool the brain (Science, 5 November 1999, p. 1071). Tiny tracks of long-gone blood vessels “can be difficult to sort out in CT scans, yet that's the only way to get the information,” he says. A CT scan through the snout of the giant sauropod Diplodocus revealed rows of teeth ready to emerge—something that had never been seen before. Such glimpses of biological infrastructure help paleontologists identify new, measurable characteristics that they can use in sorting out evolutionary connections among creatures. Skulls, with their innumerable recesses and canals that conduct nerves and blood vessels, are particularly rich sources of evolutionary information. The ease of transmitting CT data will also make important, often inaccessible, specimens widely available. Scanning programs are already under way in various countries. Fred Spoor, an anthropologist at University College London, is now working with the National Museums of Kenya in Nairobi to help scan their fossils. The images will serve as an archive and as easily transferred digital “casts” of the specimens. Washington University's Conroy and Horst Seidler of the University of Vienna have similar collaborations in Tanzania, Ethiopia, and South Africa. In the United States, one center of paleontological CT is at the University of Texas, Austin. In 1997, Rowe and colleagues William Carlson and John Kappelman set up a state-of-the-art scanning lab. After securing$1.5 million in funding, Rowe and his team spent a year designing the specifications for two industrial scanners: a high-power source to probe objects as dense as meteorites, and a low-power source to examine fine detail.

Since the lab opened for business 3 years ago, more than 100 scientists have had their fossils probed and archived on CD-ROMs. Rowe's team provides advice on how to scan fossils and can turn the massive data files into a 3D image. Scanning prices begin at $104 an hour; and because the National Science Foundation helps support the facility, researchers with NSF grants get a 50% discount. Rowe says he has clients in the imaging lab about every week. Ultimately, Rowe would like to post on a Web site all of the several hundred specimens that have been scanned in the lab—although he promises to honor requests to wait until after publication. Digital dissection For all its benefits, even its most ardent enthusiasts admit that paleontological CT scanning is not a magic wand. “There's a wide perception that you can feed a dinosaur skull into a CT scanner and all will be revealed,” Witmer says. “It's not that simple.” If a fossil is very dense or large, for example, x-rays have trouble passing through it and the signal may be quite noisy. The type of matrix matters, too; a calcium-rich bone encased in calcium carbonate-rich sediment may appear as nothing more than hazy shadows. As a result, trading the preparator's dental pick for the image processor's computer mouse may not mean a decrease in workload. Horner has spent years trying to decipher fuzzy images and decide where to draw boundaries between rock and bone. Once the bone is located, the most challenging work has just begun. Witmer points out that physicians need years of training to read x-rays of human anatomy. Interpreting scans of extinct animals is far more difficult, especially if the specimens have been damaged after death or during fossilization. All this means that interpreting CT images demands a tremendous amount of time. “You can make many images in an afternoon. Actually working on the images takes forever,” notes Spoor, who uses CT scans to study details of hominid ear bones. “Ultimately, it will remain quite a specialized operation.” That's likely to remain true even as computers grow faster and software for image processing and data analysis ever more sophisticated. In the end, “it's not going to be the technology that provides the insight,” Witmer cautions. “It comes down to humans who can understand the complicated and voluminous information that comes out of the scanner.” 11. PALEONTOLOGY # Fossils Made to Order, Any Size 1. Erik Stokstad Large dinosaur bones are “damn cumbersome,” grumps Rolf Johnson, a paleontologist at the Milwaukee Public Museum. To figure out the posture and gait of ceratopsian dinosaurs, Johnson must examine the way their limbs articulate. But that's tough to do when a single meter-long shoulderblade can weigh as much as a bag of cement. In the early 1990s, Johnson discovered a partial solution by making fiberglass casts of Torosaurus shoulder, leg, and foot bones, inserting universal joints, and rigging them with rubber bands in a 2-meter-tall wooden gantry. It worked well—except when Johnson had to wrestle the contraption into his pickup truck and drive it to meetings. Today, thanks to industrial scanning and rapid prototyping, Johnson has it easy. His 1:6 scale replica of a Triceratops forelimb fits in a briefcase. The technology, normally used to make mock-ups ranging from juice squeezers to engine parts, can reproduce a fossil and change its scale in the process. A few massive dinosaur bones have already been shrunk to a more manageable size, and in principle tiny bones could just as easily be enlarged. What's more, the right-sized replicas can be mailed to collaborators at a reasonable price—try shipping an 8-meter-long Triceratops—and then stored on a shelf. One of the first attempts to scale down a fossil came in 1998, when Jeff Wilson, a graduate student at the University of Michigan, Ann Arbor, teamed up with Arthur Andersen of Virtual Surfaces Inc. in Mount Prospect, Illinois, a company that edits digital versions of industrial parts. They converted a meter-long, 115-kilogram humerus of a sauropod into a version that was just 12 centimeters long, yet completely realistic to the eye. “Basically everything you could see on the sauropod, you could see on the small bone,” says Hans Larsson, a graduate student at the University of Chicago. “Now you can store a 60-foot [18-meter] sauropod in a filing cabinet.” Johnson's miniature dinosaur originated with the Smithsonian Institution's 1998 decision to disassemble its Triceratops, on display since 1905 and badly in need of conservation. While the bones were accessible, Ralph Chapman, a paleontologist and morphometrician at the Smithsonian's National Museum of Natural History in Washington, D.C., decided to create a digital replica. He struck up a collaboration with Andersen and Lisa Federici of Scansite, a company in Woodacre, California. Once they had a digital rendition, Federici persuaded toy manufacturer Hasbro to create a scaled-down physical model in its Cincinnati facility. The technique, called stereolithography, uses a laser to cure light-sensitive resin. Layer by layer, the tiny Triceratops bones rose from a vat of liquid resin. The model is much more than a plaything, and it provides unique insights into the specimen and the ancient animal's physiology. While examining the Triceratops bones with Chapman, Johnson realized that the front shoulder blade seemed too small for the end of the humerus. The error occurred because the skeleton is a composite, assembled from at least 10 individuals. And the mismatched joint might have prevented the paleontologists from knowing the true extent of limb motion. “I don't think I would have noticed that with the real fossils, because we couldn't put them together and move them,” Johnson says. Another insight came during a visit to the Smithsonian by Kent Stevens, a paleontologist and computer scientist at Oregon State University in Corvallis. To make the elbow more realistic, Stevens and Chapman cut up a computer mouse pad and added it as faux cartilage. When they rotated the ulna, the pair noticed that Triceratops's elbow could have locked in place. Hoofed animals such as cows and horses lock their limbs in a similar way to sleep standing up. Triceratops may have snoozed while upright too, or it might have braced itself while locking horns. “We would never have gotten into that idea if we had been solely looking at 3D software,” Stevens says. Both Stevens and Johnson say that handling the bones provides a crucial reality check for computer models of locomotion. “I'm suspicious of only scanning the bones and playing with them in the computer,” Johnson says. “You can make the computer do things that may look realistic but in fact are not biomechanically reasonable.” Prototyping makes it possible to constrain speculation about the awkwardly heavy bones of large beasts. The technique also lets paleontologists touch parts of a fossil that would otherwise be inaccessible without sawing. In 1995, Larsson and Andersen created a cast of the braincase of a predatory dinosaur called Carcharodontosaurus. After scanning the 1.6-meter-long skull by computed tomography (CT), they used a prototyping technique that lays down sheets of paper and glue, then trims them to size with a laser. Right now the technique is probably too expensive for most paleontologists. A resin prototype of a 10-centimeter-long bone goes for about$1000, Andersen says, while Chapman's pro bono Triceratops would have cost upward of $200,000. “Rapid prototyping is something we're all waiting for,” says Mark Norell, a paleontologist at the American Museum of Natural History in New York City. He's optimistic that the technology will become much cheaper and provide even higher resolution. Indeed, prototypes consisting of cornstarch and sugar are already being made at a cost of about a dollar per cubic inch. And desktop prototyping machines that resemble inkjet printers are on the horizon. Soon, instead of shipping bones by mail, paleontologists might simply FTP data from a CT scan and have their colleagues print out 3D replicas. 12. PALEONTOLOGY # CT Sleuthing Uncovers Fossil Misfits 1. Erik Stokstad Useful as they are for probing subtle features of fossil specimens, computed tomography (CT) scans are equally adept at singling out parts that don't belong. At the University of Texas's CT lab, Tim Rowe has identified mismatches that fooled even expert eyes. Last July, artist Stephen Czerkas of The Dinosaur Museum in Blanding, Utah, brought the now-infamous Archaeoraptor fossil to Rowe's lab in Austin. Czerkas hoped to get valuable information about bones still encased in rock. Instead, much to his disappointment, the scanner revealed cracks suggesting that the body and tail of the fossil had come from different animals (Science, 14 April, p. 238). “It was a shock to see this,” says Rowe, whose findings are being reviewed for publication. In January, a biologist who wishes to remain anonymous arrived with a small primate skull to be scanned. He had bought it from a fossil dealer for$2600, thinking it would make a good research specimen. After five CT slices, Rowe realized the entire skull was a clever fake carved out of dental amalgam, although it contained jaw fragments and some real teeth. The biologist was mortified, Rowe says. “He wouldn't even touch the specimen again.”

Rowe notes that CT scanning could be a valuable tool for auction houses, museums, or anyone else interested in authenticating fossils. As in the case of the biologist, money as well as knowledge may be at stake. People who inadvertently donate bogus fossils to museums, for example, may see their tax deductions plummet. “I think there are going to be a lot of victims as these forensic techniques are put into play,” Rowe says.

CT scanning can also reveal less fraudulent artistry. Early preparators of dinosaurs were highly skilled at reconstructing missing or damaged bone, but their anatomical interpretations can mislead present-day paleontologists if they remain undetected. “Very often the fakery—designed to make something look more presentable—is difficult to detect,” says Larry Witmer of Ohio University College of Osteopathic Medicine in Athens. In the past, some fossil experts have missed the plaster and described it as bone. Adds Jack Horner, a paleontologist at the Museum of the Rockies in Bozeman, Montana, “CT scanning solves that problem.”

13. SOCIAL SCIENCE

# Stress: The Invisible Hand in Eastern Europe's Death Rates

1. Richard Stone

The end of communism opened up a life of economic uncertainty in the Eastern Bloc. And that, say scientists, may be exerting a deadly effect on residents

BudapestSoon after the former Eastern Bloc nations tossed off communist rule in the late 1980s and the Soviet Union imploded, people throughout Eastern Europe began dying in droves. Life expectancy plummeted. By 1994, for example, reaching the age of 57 was enough to put Russian men on the right side of the Bell curve. Even more frightening are the demographics: The groups experiencing the highest rates of premature death are young and middle-aged men. Traditional risk factors such as bad diet, smoking, excessive alcohol consumption, and infectious diseases all claim a share of the rising mortality in this part of the world, but they can't explain the growing disparity in life expectancy between East and West, researchers say. So what could be preying on a generation that should be in prime health?

On one level, the main culprit is clear: coronary heart disease. “What's killing them is diseases of the heart,” says Gerdi Weidner, a psychologist at the State University of New York, Stony Brook. But Weidner's diagnosis—offered to a select group of 40 scientists from a range of disciplines at a NATO workshop convened here from 21 to 23 May to discuss Eastern Europe's epidemic of heart disease—wasn't based on physical symptoms alone. She and other presenters made the case that many Eastern Europeans may be dying from broken hearts. “The key words are ‘giving up,’” says conference co-director Maria Kopp, a behavioral scientist at Semmelweis University in Budapest. When Eastern Europeans gained their freedom more than a decade ago, Kopp says, “people had very high expectations” that their lives would improve. For many, those hopes were dashed quickly by the bumpy transition to a market economy. Disillusionment led to stress and depression. And depression was a harbinger of death.

Gender gap.

In many Eastern European countries, a chasm in life expectancy has opened up between men and women.

View this table:

For gene jocks, that may be hard to swallow. “That social change can affect health is a fairly novel idea to a lot of biomedical scientists,” says demographer Virginia Cain of the Office of Behavioral and Social Sciences Research at the U.S. National Institutes of Health (NIH). But for social scientists the hypothesis is in vogue. For instance, the European Science Foundation has just launched a 4-year project involving some 50 scientists to probe the link between psychology and mortality in Eastern Europe. Such a major research effort makes sense, says Cain: “The change to market economies provides a natural experiment to look at the impact of rapid social change on health.”

Death behind the Iron Curtain.

A gap in life expectancy between Eastern and Western Europe opened up more than half a century ago in the aftermath of World War II. “In Eastern Europe, you had a disastrous transition from one type of mortality to another, from infectious diseases to noncommunicable diseases,” explains epidemiologist Martin Bobak of University College, London. Life expectancy stagnated in Eastern Europe until the late 1980s, apart from an uptick in the Soviet Union around 1985 in the wake of Mikhail Gorbachev's short-lived antialcohol campaign, says Vladimir Shkolnikov of the Center for Demography and Human Ecology in Moscow. In the meantime, Westerners, eating better and exercising more, were living longer with each passing year.

In 1989, Poland, Hungary, and Czechoslovakia all overthrew their oppressive communist regimes, and other Eastern European countries began following suit. Euphoria, however, soon gave way to uncertainty. People were in control of their own lives, but life was like walking a tightrope with no social safety net. Death rates skyrocketed and life expectancy plummeted, bottoming out about 6 years ago depending on the country. “The crisis goes along with the relative success of the transition to a capitalistic society,” says Clyde Hertzman, an epidemiologist at the University of British Columbia in Canada. “Countries like Russia are relative basket cases.”

Researchers seeking to unravel this trend discovered a multitude of causes. Smoking was the culprit in some countries, while poor diets—a lack of fruits and vegetables—led the way in others. Studies showed that the region's health care systems, while frayed, are not to blame for the life-expectancy gap, says Margareta Kristenson of Linköping University in Sweden. Nor is pollution the answer: Industrial emissions fell sharply in the late 1980s and early 1990s as state-controlled factories floundered.

The Baltic blues.

Some of the most telling studies have compared the former Soviet countries of Estonia and Lithuania with Sweden, across the Baltic Sea. Margus Viigimaa of Tartu University Hospital in Estonia and his colleagues, for instance, sought to understand why Estonians are on average three times more likely than Swedes to die of coronary heart disease. They examined 274 Estonian and 271 Swedish men and women, aged either 35 or 55, and had them fill out questionnaires. Diet didn't appear to be the key to the puzzle: The researchers found almost no difference between the groups in total cholesterol and triglycerides or body fat, although the Estonians generally had lower levels of “good” cholesterol, high density lipoprotein.

The big behavioral difference, they concluded, was smoking habits. Half the Estonian men aged 35, for example, were smokers, more than three times as many as in Sweden. (To tackle this scourge, the Estonian government in 1998 instituted a total ban on tobacco advertising—the only Eastern European country so far to do so.) But even this striking difference could not explain the sharp disparity in heart disease rates, so the researchers had to look elsewhere for clues. What they have found, Viigimaa says, is that “psychosocial stress is very important.”

Following this up, psychophysiologist Sarah Knox of the NIH's National Heart, Lung, and Blood Institute in Bethesda, Maryland, and her Swedish colleagues found that in general, the Swedes reported feeling more self-confident, more in control of their lives, less depressed, and enjoying a higher quality of life than their Estonian counterparts. And they found some corroborating physiological evidence. After accounting for confounding variables from diet to health care, Knox told the meeting, the younger set of Estonian women in the group Viigimaa studied—the 35-year-olds—who report feeling “less valued” tended to have higher heart rates. A chronically elevated pulse increases the risk of damage to the endothelium of blood vessels, Knox says. And such injury is the first step toward atherosclerotic lesion formation, she says.

A similar study comparing Swedish and Lithuanian men points to psychosocial stress as a contributor to the fourfold higher heart disease risk in Lithuania. A group led by Kristenson and Zita Kucinskienë of Vilnius University found that Lithuanians in Vilnius, by their own measure, feel less in control and more exhausted and depressed than Swedes in Linköping. Kristenson's group devised a lab stress test in which subjects were asked to recall a disturbing event. The Swedes, they found, had a stronger biological reaction to the stress: Blood levels of two steroid hormones—cortisol and prolactin—shot up during 6 minutes of brooding. The Lithuanians, meanwhile, had higher baseline levels of the stress hormones and showed an attenuated response—suggesting that their bodies are perpetually stressed. “When you're chronically stressed, you're more likely to give up on life,” suggests Kristenson.

That jibes with recent findings tying psychosocial stress to coronary heart disease in Western Europe and the United States. Two years ago, for example, the Whitehall II study reported that British civil servants in low-level jobs who claimed to work hard with little reward were twice as likely to develop coronary heart disease as were higher ranking civil servants with more job satisfaction. (Both groups of civil servants had similar biomedical profiles, suggesting that diet, at least, did not account for the difference in heart disease rate.) And a study published in the 8 May issue of the Archives of Internal Medicine links depression to elevated risk of heart disease. After analyzing data on 7893 men and women enrolled in a major U.S. nutrition study, a group led by Amy Ferketich of Ohio State University School of Public Health in Columbus reported that people rated as depressed were 70% more likely than nondepressed individuals to develop coronary heart disease.

Like losing a loved one?

One researcher at the meeting compared the stress felt by Eastern European men to the loss of a spouse. Many studies have shown that widowers are more likely to die soon after the loss of their wives than are widows, says psychologist Camille Wortman of the State University of New York, Stony Brook, and widowers aged 25 to 64 are most vulnerable to an early death—from suicide, alcohol-related illness, accidents, or heart disease. Men tend to cope more poorly than women with the sudden loss of a spouse, she says, and “it's always the young men who are most at risk.” A similar pattern shows up in Eastern Europe, points out Wortman, who suggests that there's something “toxic” about a strong violation of expectations. Other meeting participants said they are impressed with this apparent connection and intend to pursue it. “This bereavement, this giving up, seems to be the most important thing,” says Kopp.

Researchers are now struggling to help people at risk, particularly young men, whom Hertzman calls “socially sensitive canaries.” Eastern European scientists need to persuade their governments to make health a top priority: “We need to convince the parties that this is not an issue for political debate,” says epidemiologist Peter Jozan of Hungary's Central Statistical Office. But it won't be easy to help residents regain a sense of control over their destiny. “You can't just tell a 15-year-old girl who's overweight and hates herself and thinks nobody loves her that she should give up smoking and lose weight,” says Kristenson. “You need to make life worth living.”

If these analyses are correct, the key to improving life expectancy in Eastern Europe lies with the region's economy. “Now that the free-for-all is over,” says Hertzman about the shift from a command to a market-driven economy, “it's time for recovery.”