# News this Week

Science  20 May 2005:
Vol. 308, Issue 5725, pp. 1096
1. CELL BIOLOGY

# Korean Team Speeds Up Creation Of Cloned Human Stem Cells

1. Gretchen Vogel

With speed and efficiency that will make waves in laboratories and legislatures around the world, scientists have created nearly a dozen new lines of human embryonic stem (ES) cells, ones that for the first time carry the genetic signature of diseased or injured patients. Last year, a group led by veterinarian Woo Suk Hwang and gynecologist Shin Yong Moon of Seoul National University reported the first—and until now the only—derivation of ES cells from human nuclear transfer experiments (Science, 12 March 2004, p. 1669). Those efforts yielded just one cell line from more than 200 tries, but the researchers report online in Science this week (www.sciencemag.org/cgi/content/abstract/1112286) that they can consistently derive a cell line in fewer than 20 tries.

The dramatic increase in efficiency suggests that creating genetically matched ES cell lines for patients needing some kind of cell transplant might not be impractical. “It's a breakthrough that I didn't think would happen for decades,” says developmental biologist Gerald Schatten of the University of Pittsburgh in Pennsylvania, an adviser to the Korean team and an author on the paper. Developmental biologist George Daley of Harvard University calls the work “spectacular.” And the work may influence the ongoing political debate over whether research with human ES cells, whether cloned or not, is ethically justified. “Some people will hate it, others will love it,” says Rudolf Jaenisch of the Massachusetts Institute of Technology. “But it puts the discussion on a very firm footing now. People will have to rethink the argument that it's not efficient.”

The new ES cell lines were created by replacing an oocyte's nucleus with one from a somatic cell and then chemically kick-starting development of the egg. Scientists similarly created Dolly the sheep in 1996 and since then have used nuclear transfer to clone thousands of cattle, mice, and other animals. Hwang and his colleagues had no intention of cloning a person, however. They only allowed the human embryos to develop for 6 days, just long enough to derive stem cells that, in theory, can form any cell type in the body.

One important factor in his team's success, Hwang says, was the use of freshly harvested oocytes from fertile women instead of ones left over from fertility treatments. The age of donors may also be key. Whereas oocytes from women in their 30s yielded on average one ES cell line for every 30 tries, those from younger donors yielded one line for every 13 tries. In nine cases, it took only a single donation of oocytes from a woman to produce a new line. (Each donation yields about 10 oocytes.)

The Korean team developed several techniques to improve their efficiency. For example, instead of using a needle to suck out the egg's nucleus, they make a small tear in the egg and gently squeeze out the chromosomes. They then insert a skin cell through the tear and apply an electric shock to fuse the two cells.

Most ES cells are derived by applying antibodies to a blastocyst-stage embryo that kill its outer cell layer and leave the inner cell mass. Hwang, Moon, and their colleagues simply put a blastocyst on a layer of human feeder cells and found that the blastocysts naturally formed colonies of ES cells. They exhibited key markers of ES cells and could form skin, muscle, and bone cells, among others.

Last year, because they had used a cell from the ovary of the oocyte donor as the nucleus donor, the Korean team could not rule out that the ES cell line was the result of parthenogenesis: an unfertilized egg starting to divide on its own. This time, except for one line, the oocyte and skin cell donors were different. In all 11 cases, the genetic fingerprint of each line matched that of the skin cell donor.

Nine of the 11 cell lines are derived from people, ranging in age from 10 to 56, who have suffered spinal cord injuries. The team has begun to test some of the lines in animal models of spinal cord injury, but Hwang cautions that they remain years away from transplanting the cells into people. “We have to be overconvinced” that the cells are safe, he says.

Another line is derived from a 2-year-old boy who has congenital hypogammaglobulonemia, a genetic immune deficiency. In theory, scientists could correct the genetic defect in the stem cells and then reinject them into the boy. Indeed, Jaenisch, Daley, and their colleagues have used such a strategy to treat mice with a similar genetic defect. Nevertheless, Hwang stresses that the boy's parents and the spinal cord patients were explicitly told that the team's research was unlikely to help them directly—even though the informed consent form used was, by Korean law, mandated to suggest such a possibility.

Although also unlikely to be employed for treatment, another ES cell line, derived from a 6-year-old type 1 diabetes patient, should interest scientists. “The possibility of being able to study disease in a culture dish is very exciting,” says Douglas Melton of Harvard University, who has recently received permission from the school's ethics committee to derive ES cells from diabetes patients. “If we could make T cells and β cells in a dish—we're not there yet, but we're getting closer—then we could compare the diabetic cells to wild-type cells and ask what goes wrong,” he explains. “For the first time we will have a chance to study the root causes of the disease.”

The improved skills of the Korean group nevertheless raise difficult ethical questions (see www.sciencemag.org/cgi/content/abstract/1114454). For example, there may be increased demand among scientists for fresh oocytes from young, fertile women. Oocyte donation is usually a safe surgical procedure, but serious complications can arise. The hormones given to trigger production of extra eggs can also cause vomiting, headaches, mood swings, and hot flashes, and the long-term consequences of superovulation aren't well understood.

In the United States, a National Academies panel recently recommended that donors of oocytes should not be paid (Science, 29 April, p. 611). In Korea, the researchers were allowed to cover travel costs of donors, but they say that no one requested reimbursement and that no payments were made.

Bioethicist Norm Fost of the University of Wisconsin, Madison, says the team's efforts to inform oocyte and cell donors was sound, but he questions using children as skin cell donors. “The [skin biopsy] that they're doing is of almost no risk and trivial discomfort,” he says. “But the default position is that when you're doing nontherapeutic research, you should use adults first.”

The new results may heat up the political debate over human ES cells. Congress is expected to vote on expanded funding for ES cell research this summer, and in Massachusetts, home to Melton's group, Governor Mitt Romney has said he will veto a new law that would specifically allow human nuclear transfer experiments. (The legislature is expected to override the veto.)

Jaenisch notes that the Koreans' successes don't change the poor odds of cloning a person: As animal cloners have found, only a tiny percentage of blastocysts develop to term when implanted in surrogate mothers. “Reproductive cloning is not safe, and it will not work,” he says. Most scientists agree, but given the unregulated nature of many infertility clinics, that may not be enough to stop renegade doctors from trying. What is certain, however, is that the new results will accelerate the already-racing stem cell field.

2. COLLABORATIONS

# Japan Bars Indian Physicists From Lab

1. Dennis Normile

TOKYO—Several Indian physicists have been blocked from visiting a Japanese research lab in the past year because of what appears to be an overzealous interpretation of rules aimed at restricting the spread of nuclear weapons. Two Japanese ministries are at odds over the unofficial policy shift, which is slowing research and raising questions about future collaborations between the two countries.

The snafu mostly involves visas for Indian scientists hoping to work at Japan's High Energy Accelerator Research Organization (KEK) in Tsukuba, although there are reports of problems visiting other labs. KEK is the site of Belle, a 13-nation experiment to explore why the universe has more matter than antimatter. Last May, after making two trips to KEK, graduate student Garima Gokhroo of the Tata Institute of Fundamental Research in Mumbai learned that her visa application had been rejected. Over the next several months, at least one Tata colleague and at least three researchers from Punjab University in Chandigarh were also denied visas to visit KEK. Their plight has recently come to light.

The Indian scientists say they were never given a reason for their rejections, and Masanori Yamauchi, a KEK physicist and spokesperson for the Belle collaboration, says he has been unable to get an explanation from Japan's Ministry of Foreign Affairs, which decides on visas. Contacted by Science, a spokesperson for the ministry declined to describe the criteria for granting or denying visas or say if Indian physicists are receiving special scrutiny.

But an official at the Ministry of Education, which recently started its own investigation, says the problem stems from concerns that India has declined to sign the Nuclear Non-Proliferation Treaty and other agreements intended to control the flow of sensitive weapons technologies. A change in personnel in the visa office has apparently resulted in a new hard line.

Yamauchi emphasizes that there is no connection between weapons technologies and the particle physics being studied at KEK. Tariq Aziz, a physicist in Tata's Department of High Energy Physics, notes that scientists visit Europe's high-energy physics lab CERN and the U.S. Fermi National Accelerator Laboratory “with no problem.”

Japan's Education Ministry, which sponsors KEK and is promoting greater scientific cooperation across Asia, is embarrassed by the flap. “We are struggling to get appropriate visas for Indian scientists,” says the official, who did not want to be identified. He says there could be a resolution “soon.”

KEK's Yamauchi says that the 400-member Belle collaboration can continue without its Indian colleagues, but their absence is hurting data analysis from the experiments. “We are suffering there,” he says. Tata's Gokhroo says her doctoral work “has definitely been delayed.”

There could also be long-term consequences, including Tata's ability to play a role on a Belle upgrade and on the proposed Next Linear Collider. “I won't be able to ask for funding if our researchers aren't going to get visas,” says Aziz.

3. NANOTECHNOLOGY

# Color-Changing Nanoparticles Offer A Golden Ruler for Molecules

1. Robert F. Service

ANAHEIM, CALIFORNIA—For researchers looking to monitor the nanoscale movement of biomolecules, good techniques are hard to come by. One that's been widely popular among biologists is to tag molecules of interest with different fluorescent dyes and hit them with a burst of light. Because of the way the dyes absorb and re-emit each other's light, tags very close together glow a different color from those farther apart. Unfortunately, the technique—known as fluorescent resonance energy transfer (FRET)—works only if the dye molecules are less than 10 nanometers apart, and the tags typically wink out after less than a minute of light exposure. Now a group at the University of California (UC), Berkeley, has come up with a novel molecular ruler that solves both problems at once.

At a meeting* here last week, UC Berkeley chemist Paul Alivisatos reported a way to use pairs of gold nanoparticles to measure distances out to 70 nanometers and keep track of their targets indefinitely. “It's really cool,” says Thomas Kipps, a cancer cell biologist at UC San Diego. By extending the length of the ruler, Kipps says, nanoparticles—also known as quantum dots—offer the opportunity to gauge the proximity of molecules across stretches equivalent to large complexes of proteins. That, in turn, may make it possible to track events from the binding of DNA strands to one another to the ability of proteins called transcription factors to bind with and initiate genetic transcription.

Gold nanoparticles have been used for sensing since 1997, when a team at Northwestern University developed a scheme for detecting specific snippets of DNA with a simple color-change test. The researchers attached gold nanoparticles to single-stranded DNAs designed to home in on target DNA sequences. As the DNA strands bound to their quarries, they pulled the gold particles together tightly enough to change the way their electrons moved—a property known as plasmon resonance. The shifting electronic behavior altered the wavelengths of light the particles scattered, changing their color (Science, 22 August 1997, p. 1036). The experiment used hordes of nanoparticles to create a color change that was visible to the naked eye. But other studies suggested that even two particles should produce a shift visible through a microscope.

Alivisatos, Carsten Sönnichsen, Björn Reinhard, and Jan Liphardt—all colleagues at UC Berkeley and Lawrence Berkeley National Laboratory—decided to see for themselves. First, using a pair of proteins as molecular glue, they bound 40-nanometer gold nanoparticles to a glass slide. When they shined white light on the slide, far-apart particles scattered green light most strongly, with wavelengths of about 540 nanometers; light scattering from particles closer together shifted to the red end of the spectrum by about 20 nanometers. Between those two extremes, wavelength changed steadily with distance.

With their new molecular ruler in hand, Alivisatos and his colleagues set out to track the binding and unbinding of DNA. They started with a solution of pairs of gold nanoparticles tethered by snippets of single-stranded DNA. Under white light, the particles scattered light at about 550 nanometers. The researchers then added DNA strands that were complementary to the tethers. The newly introduced DNA strands bound to the tethers, stiffening them enough to push the nanoparticles apart by about 2 nanometers. As that happened, the wavelength of light scattered by the nanoparticles shifted toward the blue end of the spectrum by a few nanometers.

Alivisatos says he hopes the longer-lived, longer-range nanoparticle-based FRET will eventually overtake its organic cousin for measurements in which background light-scattering is low. That is already happening in the world of quantum dots, he notes, in which tiny inorganic nanoparticles are beginning to replace organic fluorescent dyes in a wide range of applications.

• *NSTI Nanotech 2005, Anaheim, California, 8–12 May 2005.

4. NASA BUDGET

# Griffin Names Winners and Losers in Cost Squeeze

1. Andrew Lawler

Declaring that NASA “can't afford to do everything on its plate,” the agency's new chief last week laid out sweeping changes to the U.S. civilian space program—including the $5.5 billion science program. Michael Griffin says he plans to scale back space station research, defer work on a future Mars robotic mission, inject more cash into NASA's struggling earth science effort and servicing and safe deorbiting of the Hubble Space Telescope, and back a mission to Jupiter's moon Europa using a conventional rather than nuclear system. He also pledged to protect the science budget from the cost of sending humans to the moon and Mars. Griffin's plans for the$16.2 billion agency were laid out in a bulky budget document for the current year sent to Congress 11 May and reinforced at a Senate hearing the next day. “We have tried to be sensitive to the priorities of the affected research communities and have listened carefully to their input,” he wrote.

Griffin was blunt about NASA's fiscal crisis, which includes $500 million in overruns on projects from the Mars Reconnaissance Orbiter to the Pluto mission slated for launch next year, more than$400 million in congressional pork, and the increased costs to get the space shuttle flying again. The new operating plan shaves $53 million this year from a$4 billion space and earth sciences budget. Bigger savings, he says, would come from deferring work on human exploration technologies, reducing the number of contractors involved in building a new human exploration vehicle, and scaling back the Prometheus nuclear system championed by his predecessor Sean O'Keefe (Science, 30 January 2004, p. 614).

Spending $270 million rather than$431 million this year on the system would torpedo plans for a probe to Jupiter to examine that planet's array of icy moons. But Griffin, appearing for the first time before the new Senate Appropriations Subcommittee on Commerce, Justice, and Science, assured legislators that a mission to Europa “remains a very high priority” and promised a detailed plan for a flight using standard chemical propulsion. Cost overruns killed an earlier proposed Europa mission.

Although overall spending on space and earth sciences remains largely unchanged, the shifts within that budget have big implications for individual projects. For example, NASA intends to defer launch of the Mars Science Laboratory from 2009 to 2011 and scale back funding for the Space Interferometry Mission and the Terrestrial Planet Finder—two missions slated for launch later in the next decade and designed to seek extrasolar planets. Some of that money would be diverted to earth science, and another portion would be used to ensure potential Hubble servicing and, eventually, a safe deorbiting of the massive telescope. “We have heard the response of the science community, and we in turn are being responsive,” Griffin said. A final decision on a Hubble servicing mission is expected after the second shuttle mission, now slated for September.

Griffin also suggested “alternative configurations” that would allow NASA to complete the space station with fewer than the 28 shuttle flights now planned. “Some of the research [to be done] on the utilization flights could be deferred,” he suggested. NASA's operating plan cuts $106 million from the$1-billion-a-year biological and physical research effort and assigns a lower priority to basic research using organisms such as cells and rats, as well as fundamental research with no link to human exploration. “Research [on the station] is valuable and must be done,” Griffin said, “but if it is delayed a very few years … then … that delay would be worth it.” He also promised legislators that exploration would trump the overall science budget only “under the most extreme budget pressure.”

Despite their concerns about individual projects, legislators seemed to welcome Griffin's direct approach to the agency's fiscal troubles. “Some of the things you've said give us heartburn,” said Senator Barbara Mikulski (D-MD). But “thank you for your candor.” Given the difficult choices Griffin must make, that is high praise.

5. U.S. NUCLEAR WEAPONS

# Bunker Buster Shot Down in Opening Volley

1. Eli Kintisch

Opponents of nuclear “bunker buster” weapons have scored a victory in the first round of the annual fight over U.S. nuclear policy.

One beneficiary would be the U.S. Army Medical Research Institute of Infectious Diseases (USAMRIID) at Fort Detrick, Maryland. Already slated for a $1 billion facilities expansion in the president's 2006 budget request, USAMRIID stands to gain staff from other facilities and join a DOD center of excellence in biodefense. Five other joint Army, Navy, and Air Force centers will study topics such as chemical defense and infectious diseases. The pathology institute, with a current 820-member staff that includes about 120 scientists, would, however, get lost in the shuffle. It began in 1862 as a museum for specimens from Civil War casualties. In 1946, Congress created AFIP, which specializes in diagnosing difficult disease cases for both military and civilian doctors. Its experts were “among the giants,” and educational training there was “legend,” says pathologist Fred Gorstein of Thomas Jefferson University in Philadelphia. Recently, AFIP scientists fingered the virus that caused the 1918 pandemic influenza, identified victims of the 9/11 terrorist attacks, and helped investigate the 2001 anthrax poisonings. Staffers have known for a few years, however, that DOD might close AFIP as part of efforts to eliminate civilian services, and some have moved on. Under the plan, only AFIP's renowned tissue repository and the flagship National Museum of Health and Medicine, with its displays on Civil War medicine and preserved body parts and fetuses, will remain. Diagnostic pathology tasks will be outsourced, and DOD will shift AFIP's work on a DNA registry and forensics to Dover Air Force Base in Delaware. Several AFIP scientists declined to comment. But pathologist William Travis, who left in January for Memorial Sloan-Kettering Cancer Center in New York City, called it “a tragedy” to close “a national medical treasure.” He would like to see Congress rescue the institute, possibly by contracting it out to the American Registry of Pathology, a nonprofit organization chartered by Congress that links AFIP to civilians. Travis also worries about the fate of the tissue repository, which includes unique specimens of rare tumors and infectious diseases. “It loses its value if separated from pathology expertise,” he says. The base closure plan must be approved by an independent, nine-member commission and then by Congress, which is not allowed to tinker with its recommendations. 7. HIGHER EDUCATION # Harvard Pledges$50 Million To Boost Diversity on Campus

1. Andrew Lawler

Harvard University plans to spend at least \$50 million over the next decade to create a more diverse academic community in all disciplines, including throughout the sciences. President Lawrence Summers announced the outlay this week after receiving two reports commissioned in February following his comments about the ability of women to do science, which triggered a national debate.

The initiative will tackle all aspects of gender and minority issues, from the safety of women working late at night at research labs to the need for a high-level advocate within the Harvard administration. Such a comprehensive strategy is essential, say the chairs of the two task forces that reported to Summers. “Women need to see careers in science as desirable and realistic life choices,” says Barbara Grosz, a computer scientist who led one of the task forces that focused on science and engineering. A second task force, led by science historian Evelynn Hammonds, examined challenges facing all women faculty.

Outside researchers are impressed with the breadth of the recommendations. “This is very encouraging,” says Donna Nelson, a chemist at the University of Oklahoma, Norman, who tracks the status of women and minority academic scientists. “If they can implement this, they can take a leadership role.”

Harvard has long been criticized for its lack of diversity of science faculty in several disciplines, a situation made worse by Harvard's decentralized structure and its policy not to grant tenure to junior faculty, task force members said. Last year, for example, four women and 28 men in the school of arts and sciences received tenure offers. But the long-simmering issue did not come to a head until Summers's comments at a January workshop on women in science became public (Science, 28 January, p. 492). The resulting outcry triggered a faculty vote of no confidence in Summers, who apologized repeatedly.

Hammonds's committee called for a senior provost for diversity and faculty development to work with Harvard deans to promote gender and ethnic equity. Harvard Provost Steven Hyman hopes to name that person—who likely would come from within Harvard—by September. The panel also proposed two funds, one to provide partial salary support for hiring scholars who increase diversity, the second to fund their labs. It said Harvard should begin to gather systematic data on faculty hiring, retention, and other measures and make the academic culture more family-friendly, through enhanced maternity leave practices, child-care support, and adjustments to the tenure clock. Grosz's panel urged the university to set up summer research programs for undergraduates, expand mentoring for all students, and provide research money for faculty juggling family and career.

Funding will not be a problem, Summers assured reporters, referring to the likelihood of “more resources allotted down the road.” The biggest challenge Harvard faces, he said, is to overcome “issues of culture” within a university created “by men for men.” Harvard is accepting comments on the report through the end of June, and academics around the country will be watching closely to see how well Harvard succeeds in transforming that culture.

8. GENOMIC MEDICINE

# Gene Sequence Study Takes a Stab at Personalized Medicine

1. Elizabeth Pennisi

COLD SPRING HARBOR, NEW YORK—Since its beginning 15 years ago, the Human Genome Project was sold to the public and to Congress as a biomedical effort that would ultimately bring a person's unique DNA sequence data to bear on preventing and treating disease. Now the National Human Genome Research Institute (NHGRI), which led the U.S. public sequencing effort, is about to take a controversial step toward that goal.

At the Biology of Genomes meeting here last week, NHGRI's Eric Green announced that NHGRI will launch a pilot study in which researchers will sequence a portion of DNA from 400 seemingly healthy volunteers and try to discern each person's unique genetic risk factors for disease. They also plan to study the reactions of the volunteers to learning these results. “[NHGRI] is doing a reality check: Do people really want personalized medicine?” says Kelly Frazer, a genomicist at Perlegen Sciences in Mountain View, California.

The project, dubbed clinENCODE, promises to jump-start the transition from basic biological studies to clinical genomics, and that “is what the genome project is all about,” says Richard Wilson, director of the sequencing center at Washington University in St. Louis, Missouri. But both Bruce Roe of the University of Oklahoma, Norman, and Evan Eichler of the University of Washington, Seattle, call the study as described a “terrible idea,” in part because the sequence information from each individual may not provide much relevant biomedical information.

The 400 volunteers will donate DNA and undergo a battery of tests, including blood pressure measurements and white blood cell counts. Green and his colleagues will sequence the same 1% of each person's genome, regions that are already being intensely studied by basic researchers. Green's team plans to report back any variations spotted, including ones that may explain a person's current and future health status.

It's not clear how people will react to such results. Previous studies involving genetic testing for specific diseases have suggested that people can handle bad health news. Still, many fear that this genetic information will lead to discrimination by employers and insurance companies.

Many genome scientists argue that clin-ENCODE is not the best way to explore the future of personalized medicine. “There are so many genes whose function and link to disease is unknown that the information we are going to give is of dubious nature” and may overwhelm the participants, says Frazer. If the chief goal is to test how the public reacts to personalized genome information, then why not simply do surveys or present mock sequencing results rather than incur the expense of sequencing, she and others wonder.

Even if the study provides little biomedical data, it will still be worthwhile, contends Robert Waterston, a geneticist at the University of Washington, Seattle. “We have to understand what the issues [of personalized medicine] are,” he says. “[The study] begins to challenge us to think about these things.”

9. ECOLOGY

# Biologists Find New Species of African Monkey

1. Mary Beckman*
1. Mary Beckman is a writer in southeastern Idaho.

When conservation biologist Trevor Jones peered last year through his binoculars at the shape flitting through Tanzania's Ndundulu Forest Reserve, he saw something unexpected. Out searching for a gray, pink-faced monkey called the Sanje mangabey, he instead spotted a brown, black-faced mangabey sporting an upright crest on his forehead that made the animal look “punky.” Speechless and shaking, Jones sat down. “I was gob-smacked,” says Jones, who works for the Udzungwa Mountains National Park in Tanzania.

About the same time, 350 kilometers away in Tanzania's Southern Highlands, researchers led by zoologist Tim Davenport of the Wildlife Conservation Society in Tanzania had been trying to track down an animal called the Kipunji. Local hunters often talked about the unusual monkey, but they were known to speak of spirit animals too. But real it was: Davenport first spotted the unique mangabey almost a year after his team started looking. The two groups heard about each other's findings in October, and now, on page 1161, they together describe the new species, dubbed Lophocebus kipunji.

“This is big news for Africa,” says primatologist Scott McGraw of Ohio State University, Columbus. “The chances of finding a large, noisy monkey that no one's ever [scientifically] described before makes this a rare event,” agrees primatologist John Fleagle of Stony Brook University in New York. In addition, the forest in which the teams found the species is one of the most globally significant regions for biodiversity—the now heavily threatened animal and plant species living there go back 30 million or 40 million years in history, says biologist Neil Burgess of the United Nations Development Programme and the World Wildlife Fund-USA.

Although the researchers still need a DNA sample to determine how closely related the new species is to other mangabeys, the highland mangabey looks and sounds quite different from its cousins. It utters a softer “honk-bark” compared to the louder “whoop-gobble” call of other tree-dwelling mangabeys, says Jones. Kipunji are also shy, he says, and exhibit some unusual behaviors: “Just before he flees, the male does this fantastic head- shaking behavior as if he's admonishing you.”

Mangabeys belong to two groups. One group, which includes the Sanje mangabey, wanders the forest floors and is related to mandrills. The other lives in the trees and is more closely related to baboons; L. kipunji is the third species in this group. Some researchers consider the area in which the animals were found to be the “epicenter” of baboon and mangabey evolution. McGraw hopes the kipunji will help researchers reconstruct how the two primate species, and another called the gelada, radiated out from a common progenitor.

But the new mangabey is already threatened. Preliminary estimates of its range encompass just 120 square kilometers total, and the research teams predict that no more than 500 animals exist in each forest. Davenport says conservation efforts need to be stepped up to prevent the animals in the Southern Highlands from being hunted to extinction. Burgess adds that the highland mangabey and another recently discovered shrew in the Ndundulu Forest Reserve might be the push that gets the small piece of forest rolled into Udzungwa Mountains National Park. But ultimately, he says, Tanzania is a poor country: “If we want to keep these [animals], the global community has to provide money until the country becomes a richer place.” Otherwise it may become poorer in monkeys.

10. PHYSICS

# Neutron Stars Could Test Quantum Effect

1. Charles Seife

It's one of the stranger predictions of quantum electrodynamics: In a strong magnetic field, the vacuum of space might behave like a crystal. In the 29 April issue of Physical Review Letters, Italian and French physicists argue that a peculiar star system might provide an unprecedented chance to test this prediction. “It's a really unique opportunity,” says Michael Kramer, an astrophysicist at Jodrell Bank Observatory in Manchester, U.K. “Just speculating about the possibility is very exciting.”

The excitement surrounds a binary pulsar: two neutron stars orbiting each other, discovered late in 2003. The neutron stars in the system both emit powerful beams of radiation that zap Earth at predictable intervals. This means that the system, known as J0737-3039, provides a pair of clocks to measure how the gravitational fields of the massive stars warp time and space, as the general theory of relativity predicts. Now physicists say they have found a new use for the stars.

Quantum theorists consider the vacuum to be full of particles constantly winking in and out of existence. Where those “virtual particles” encounter the powerful magnetic fields near a neutron star, light passing through should slow down and bend, just as it does inside a hunk of glass or crystal. “The index of refraction changes with magnetic field,” says Carlo Rizzo, a physicist at the Institute for Research on Atomic Systems and Complex Molecules in Toulouse, France. “And different frequencies of light have different velocities.”

Rizzo and colleagues in France and Italy hope J0737-3039 will help them measure this subtle effect. “It's like somebody in the cosmos set up a system to do the sort of experiment we want to do in the lab,” Rizzo says. At times during the stars' orbit, the beam from one pulsar passes right through the other pulsar's intense magnetic field. If x-ray astronomers observe for a long enough time, Rizzo says, they might see a bending of light rays that exceeds that due to the gravitational distortion of spacetime.

It won't be easy. “You have to observe for a long time if you want to have enough statistics,” Rizzo says. Indeed, Alice Harding, a physicist at NASA's Goddard Space Flight Center in Greenbelt, Maryland, doubts that the bending will be seen anytime soon. Not only is the phenomenon small compared to gravitational lensing, but if the neutron star's spin axis is pointing even slightly in the wrong direction, “that will wipe out any detectable effect,” she says.

But whether or not J0737-3039 is a good spot to find quantum-theoretic vacuum lensing, the system is already a laboratory for new physics. “It has almost become an industry on its own,” Kramer says.

11. ASTRONOMY

# The Hunt for Stealth Galaxies

1. Robert Irion

Astronomers are learning how to find the barest sprinklings of stars, which trace unseen pockets of dark matter in our cosmic neighborhood

If grand spiral galaxies are the photogenic pinups of astronomy, then the faint smudges of stars called dwarf galaxies are the bit players that few fans will recognize. Telescopes can barely see them, and no one knows how many dwarfs inhabit the bleak gulfs between galaxies like the Milky Way and Andromeda. But just as minor actors can steal a scene, dwarf galaxies are earning respect from astronomers who take time to stare away from the lights.

Sensitive searches of space are unveiling a growing population of “little pathetic things,” in the words of astronomer Liese van Zee of Indiana University, Bloomington. Although their stars are meager, dwarfs appear to be embedded within dense cocoons of unseen dark matter—the same mysterious stuff that composes the bulk of the universe's mass. Tracing the numbers and locations of dwarfs is giving theorists a better grasp of how dark matter has shaped the growth of larger galaxies and is revealing the smallest coherent clumps of matter within which stars can form.

Meanwhile, radio telescopes are tuning in to the faint murmurs of other small galaxies and finding huge amounts of matter—ordinary hydrogen gas—that never coalesced into stars. Some of this matter may date to the earliest history of the cosmos, giving astronomers a chance to study pristine gas unprocessed by the fires of stellar fusion. In one disputed case, researchers may have found a true “stealth galaxy,” a massive whirling disk of hydrogen that has spawned no stars at all.

These galactic shreds may have been the first substantive knots of matter to assemble in the universe. Astrophysicists think most of them collided over the eons to create big galaxies like our own—mergers that continue on a minor scale today. In that picture, today's dwarfs are the last remnants of those ancient structural seeds.

“Dwarf galaxies are our best way to figure out what the building blocks of our galaxy would have looked like,” says doctoral student Alan McConnachie of the University of Cambridge, U.K. “The dwarfs we can see are special. They are the ones that survived.”

## Imprints of tides

Astronomers must look close to home to find those survivors: Most dwarfs are too insubstantial to be seen at great distances. By studying dwarfs within our cosmic suburb, called the Local Group, astronomers can make deductions about the environs of any mature galaxy, says astronomer Michael Merrifield of the University of Nottingham, U.K.: “The immediate neighborhood of the Milky Way is a representative bit of the universe.”

The Milky Way and Andromeda dominate the Local Group, which spans about 10 million light-years of space. The group also contains a few midsize galaxies and about three dozen known dwarfs. Roughly a dozen of these dwarfs appear to orbit the Milky Way; Andromeda has a slightly bigger retinue. The rest are scattered spritzes of stars doing their own thing, with no apparent ties to the Local Group's giants.

These minigalaxies are not inert nubs. Rather, astronomers think they have histories as dynamic as those of their bigger neighbors. “They have been evolving chemically and structurally for the entire history of the universe,” says McConnachie. That evolution critically depends on a dwarf's path through space.

Most isolated dwarfs in the middle of the Local Group are “irregulars,” misshapen patches with some younger stars and rich clouds of gas. But when a dwarf approaches a big galaxy, it transforms. Hot matter on the outskirts of a massive galaxy may strip some gas out of the dwarf as it orbits, like a fresh wind clearing fog out of a city. Moreover, the big galaxy exerts gravitational tides on the dwarf. Those motions set off waves of star formation, exhausting the dwarf's remaining supply of gas—the fuel for creating new stars.

What's left is a so-called spheroidal, a barren fuzzball of old stars. These objects are ripe for cannibalism by the bullies of the Local Group, a process that astronomers can now trace from start to finish. It begins with telltale tidal distortions within the dwarfs, such as S-shaped patterns of stars in spheroidals near Andromeda. Within a few billion years, such dwarfs are doomed—sure to be dragged ever inward through the galaxy's extended halo of gas and its pervasive shroud of dark matter.

When a dwarf starts slogging through the visible outskirts of a big galaxy, gravitational tides tear it apart. Astronomers see this happening today to a stretched-out dwarf called Sagittarius, on the far side of the Milky Way. Similarly, thick streams of stars lacing around Andromeda are the sole remains of dwarfs that the galaxy recently absorbed.

Astronomers believe this process happened often in the early universe, as major galaxies assembled within dense pockets of dark matter. Dwarf galaxies have far “darker” pockets than big galaxies have, on a star-by-star basis. For instance, a nearby dwarf spheroidal called Draco may pack 200 times more mass in invisible matter than in its stars. That's an order of magnitude higher than the ratio of dark matter to luminous matter in the Milky Way. Dwarf galaxies are nuggets of dark matter, it seems, with stars sprinkled in as an afterthought.

## Calling all satellites

That blackness poses a vexing challenge. Theorists crave an accurate census of the dwarfs that populate our local cosmos to help solve a puzzle they first noted in 1999, called the “missing satellites problem.” But astronomers can't yet tell how many they've missed.

According to cosmological models, a vast web of dark matter formed in the early universe. Astrophysicists see that embryonic pattern in the subtle ripples of the cosmic microwave background, the remnant glow of the big bang itself. The web controls where and how galaxies arise. Big clumps of dark matter attract smaller ones, thanks to their powerful gravity. And ordinary matter, such as hydrogen gas, settles within the clumps and sparks the birth of stars when its density gets high enough.

But the process is messy. Small knots of dark matter can swarm for eons without merging, like leaves circling an eddy. Simulations predict that enough of these leftover dark-matter “subhalos,” as the knots are known, should fleck the Local Group to seed many hundreds of dwarfs. Where are they?

In the past 3 years, theorists have found a possible explanation: Many subhalos probably didn't stay calm enough to form stars. As hot young stars began to shine in bigger galaxies, they irradiated space with ultraviolet light. This energy would have excited hydrogen gas in the nascent dwarfs, preventing the gas from cooling enough to collapse into new stars. If stars did form, they might have wreaked havoc in the puniest subhalos. “The [gravitational] binding energies of these galaxies are so small that one supernova could disrupt the whole thing,” says Merrifield.

If this scenario is correct, surveys with optical telescopes won't find hundreds of nearby dwarfs after all. However, observers and theorists agree that some dwarfs surely await detection. For instance, new simulations by a group based at the University of Durham, U.K., point to as many as 70 visible dwarfs near the Milky Way. The galaxy's 12 known dwarfs may yet be “the tip of the iceberg,” says graduate student Noam Libeskind of the Durham team.

Astronomers are hot on the trail. Teams rely on surveys that span sweeping chunks of the sky—notably the Sloan Digital Sky Survey, based at Apache Point Observatory in Sunspot, New Mexico. The survey classifies stars by color and brightness so accurately that computers can select light from stars of a uniform type. For instance, all galaxies contain a percentage of bloated red giant stars. A surplus of such stars in a small patch of sky might trace the faint wisps of an unknown dwarf. Daniel Zucker of the Max Planck Institute for Astronomy in Heidelberg, Germany, compares this method to “finding forests by their trees.”

A team led by Zucker used Sloan data in 2004 to find a new companion to Andromeda, a barely-there dwarf called Andromeda IX. The stars are so sparse that even a detailed image from Japan's 8.2-meter Subaru Telescope in Hawaii, shown on this page for the first time, hardly reveals the galaxy. “You would have to stare at one place in the sky with a large telescope for an incredibly long time just to see some fuzziness,” Zucker says.

Andromeda IX is an anemic galaxy of old stars, according to a study led by astronomer Daniel Harbeck of the University of California, Berkeley, in the 10 April Astrophysical Journal. The dwarf must have lost its gas early on, Harbeck says, preventing a new generation of stars from forging iron and other heavy elements. And the first analysis of the motions of stars within the dwarf hints at a dominant nugget of dark matter similar to the one found in Draco, reports a team led by astronomer Scott Chapman of the California Institute of Technology in Pasadena. The fast-moving, widely spaced stars would disperse into space without the dark matter's hefty gravity. The new paper will appear in Astrophysical Journal Letters.

Less than a year after the discovery of Andromeda IX, a new Milky Way dwarf—a smattering of about 100,000 stars in Ursa Major, or the Big Dipper—took its place as the least luminous known galaxy. The new runt could keep its dubious honor for a while, says astronomer Beth Willman of New York University, whose team announced the find in March. “This object is close to the limit of what we can detect with Sloan,” Willman says, and Zucker adds that it probably wouldn't be visible with current data if it orbited Andromeda. Both Zucker and Willman see other potential dwarfs lurking in their images, but they say more-thorough surveys will be needed to find most galactic satellites and to make observations jibe with theory.

To do more, astronomers must turn a different set of eyes onto the heavens: radio telescopes. When tuned to a certain wavelength, these dishes pick up subtle signals from small galaxies—not from stars, but from gas.

That wavelength is the famous 21-centimeter line spontaneously emitted by cool, neutral hydrogen atoms. Radio astronomers have studied that line for decades, but only recently have they outfitted telescopes with the right tools to conduct broad surveys for gas-rich dwarf galaxies and other as-yet-unseen objects.

The most ambitious program is now under way at the 305-meter Arecibo radio telescope in Puerto Rico. With a sensitive new compound detector, built in Australia, the Arecibo team plans to image hydrogen emissions over 1/6 of the sky within 4 to 5 years. The survey officially began in February, but the team had already detected 165 galaxies and other objects during a commissioning run last fall.

“If a galaxy has any hydrogen gas in it, we will see it,” says radio astronomer Martha Haynes of Cornell University in Ithaca, New York, who leads the survey with Cornell colleague Riccardo Giovanelli. Among the survey's main quarries are dwarfs in or near our Local Group that have retained their gas by virtue of avoiding interactions with big galaxies.

Indiana University's van Zee found one such object serendipitously with the Very Large Array of 27 radio telescopes in Socorro, New Mexico. Among five galaxies in van Zee's study was UGC 5288, a nondescript dwarf about 16 million light-years away. Radio emissions revealed an extraordinary disk of gas, extending seven times farther into space than the galaxy's stars. “It's a huge amount of hydrogen, but it's spread out like a pancake,” says van Zee. She described the dwarf in January at a meeting of the American Astronomical Society in San Diego, California.

According to van Zee's analysis, the hydrogen is rotating peacefully. That suggests the gas was not expelled by supernovas or captured during a merger. She suspects the hydrogen is a relic of the galaxy's birth, making the disk a potentially rare sample of the gas from which galaxies arose—and relatively uncontaminated by nuclear fusion in stars.

UGC 5288 is “density-challenged,” Haynes says. “It did not have enough gravity to form in a normal way. These galaxies have a much slower process of converting their gas into stars, if at all.” Van Zee notes that UGC 5288 does contain a lot of dark matter, but some process—perhaps rapid spin at birth—spread most of its gas too diffusely.

Astronomers have long hoped to find an even more extreme object: a galaxy consisting only of gas, in which stars have never burst forth. A team led by astronomer Robert Minchin of Cardiff University in the United Kingdom made just such a claim in the 20 March Astrophysical Journal Letters. Using radio data from the Jodrell Bank Observatory in the U.K., the team found a “dark hydrogen cloud” about 1/10 as massive as the Milky Way on the margins of the heavily populated Virgo Cluster of galaxies. The starless cloud shows evidence of galaxylike rotation, Minchin says.

The observations drew worldwide attention, but few other astronomers were convinced. Nottingham's Merrifield noted that the pattern Minchin's team ascribed to a rotating disk—closely tied regions of hydrogen, some moving away from us and others toward us—could also arise from smaller blobs of gas moving in different directions. Haynes also is skeptical: “The Virgo Cluster is a tricky place to work. It's a dynamic environment,” she notes, with galaxies milling about and perhaps casting off shreds of gaseous debris.

New detailed images might settle the issue. Minchin's team used the Westerbork Synthesis Radio Telescope, an array of 14 antennas in the Netherlands, to zero in on the mystery object in late April. The team has not yet settled on an explanation for the patterns it sees. “We're working on what it means,” Minchin told Science. “It's certainly more complex than just a [dark] galaxy on its own. I still think it's a bona fide galaxy,” he says, although it may have interacted with a neighbor.

The astronomers also will use the Hubble Space Telescope later this year to scour the dark patch for hints of stars. “Watch this space,” Minchin says with a chuckle. His slogan applies equally well to those who scan the depths between giant galaxies, looking for feeble companions to help complete the tale of cosmic assembly.

12. NONPROLIFERATION

# A Radioactive Ghost Town's Improbable New Life

1. Richard Stone

The city of Pripyat, abandoned after the Chornobyl explosion 19 years ago, offers a unique trove of data for modeling a dirty bomb attack

PRIPYAT, UKRAINE—A rusted Ferris wheel groans in a stiff breeze, the only sound in Pripyat's central square. In April 1986, this attraction and the adjacent bumper cars were newly built and preparing to open for the First of May holiday. Then on 26 April, reactor number four of the Chornobyl Nuclear Power Plant exploded, spreading radionuclides across Europe. Most of the 50,000 residents of Pripyat, within eyesight of the reactor, were power plant workers and their families; everyone was evacuated. They were told to pack for a 3-day trip, but their relocation to other parts of Ukraine ended up being permanent. Nineteen years later the abandoned town is frozen in time, the dilapidated little amusement park still waiting for opening day.

In a bizarre twist brought about by the 11 September 2001 terrorist attacks, Pripyat is getting a new lease on life. People will never move back into the deteriorating Soviet-era apartments. Instead, scientists are planning to use the radioactive ghost town as a unique laboratory for modeling the dispersal of radionuclides by the detonation of a dirty bomb or an attack with chemical or biological agents. “Pripyat offers an unparalleled opportunity to fully understand the passage of radioactive debris through an urban area,” says a nonproliferation official with the U.S. State Department. Modeling in Pripyat, he says, also “can be extended to preparing us against biological and chemical aerosols.”

The surreal city's resurrection as a test bed for catastrophes gained backing at a workshop on aerosol dynamics held last month at the International Radioecology Laboratory (IRL) in nearby Slavutych, a town built to replace Pripyat. The workshop was sponsored by the U.S. Civilian Research and Development Foundation, an Arlington, Virginia-based nonprofit that funds nonproliferation efforts in the former Soviet Union. There, radioecologist Ronald Chesser of Texas Tech University in Lubbock described new models of the radioactive plumes from the burning reactor. In addition to giving a sharp picture of the accident, they can be adapted to predict the spread of aerosols in a hypothetical terrorist attack.

Two years ago, a team led by Chesser, Brenda Rodgers of West Texas A&M University in Canyon, and IRL's Mikhail Bondarkov measured radioactivity at hundreds of spots in the so-called Red Forest, a swath of dead pines west of the reactor that received lethal radiation doses from the first plume, known as the western trace. (It's called the Red Forest because the needles turned an auburn color.) Sampling 17 years after the accident, Chesser had expected a blurry approximation. “To our surprise,” he says, “we saw a very good picture of the plume” as reconstructed from particle density and deposition data: a 660-meters-wide, 290-meters-tall bell-shaped column.

Fortunately, the western trace missed Pripyat, which lies about 3 kilometers north of the reactor, but it “probably wiped out most wildlife in the Red Forest,” Chesser says. By the time the winds began pushing the plume northward, it was about half as dense, he says. To reconstruct how badly Pripyat was hit, last summer his group measured radioactivity at more than 1700 spots in and around the city. They found that the heart of the northern trace barreled just east of Pripyat (see graphic, below). If the city had absorbed a direct hit, Chesser estimates that the toll would have been roughly 6000 cancer deaths. “The winds were very, very fortunate,” he says.

The U.S. Defense Threat Reduction Agency (DTRA) intends to build on this work to forecast what would happen if a dirty bomb were to explode in a city. “We can't directly simulate this kind of attack, so we use various means to obtain representative data,” says John Pace, a meteorologist with DTRA's Chem-Bio Defense Program in Fort Belvoir, Virginia. “The advantage of Pripyat is that the radioactivity is already there.” In the city's central square, moss growing in cracks in the pavement sends Geiger counters galloping; it will be another decade before half the radiocesium deposited here will have decayed. Although Pace notes that there are “huge differences” in the consequences of a dirty bomb compared to those of the Chornobyl explosion, by focusing on the spread of material, “we can still obtain useful data that we can use to improve our capabilities to respond to urban terror attacks.”

Studying surface contamination can give clues to how aerosol deposition is affected by a town's layout, construction materials, and building positions relative to prevailing winds. “What's particularly interesting with Pripyat is that there are a number of rather tall buildings, up to 16 stories, so we can go back and gather exposure data from different levels above the ground,” Pace says. Vertical mixing of contaminants in cities, he says, “is an area where we don't have as much data as we'd like.”

Down the road, benign gases could be released in Pripyat to model dispersal. DTRA has supported similar studies. In 2001, 120 shipping containers were set up to model release scenarios in the Mock Urban Setting Test at the Dugway Proving Ground in Utah (Boundary-Layer Meteorology, June 2004, p. 363). In Oklahoma City in 2003, DTRA and the Department of Homeland Security sponsored a study in which an inert tracer gas was released downtown. A similar experiment is planned for Madison Square Garden in New York City this summer.

In living cities, however, constraints on sensor placement and on release locations and times can limit the range of data collected. “A site like Pripyat would offer more freedom in that regard,” says Jeremy Leggoe, a chemical engineer at Texas Tech who has modeled the influence of vegetation on aerosol dispersal. Pripyat has disadvantages: For example, vegetation that has gradually been engulfing the city would have to be cut back. “That's particularly important, since in a real event, a large proportion of the exposure that you're concerned about—initial victims and emergency responders—will take place at ground level,” Leggoe says.

Faced with such obstacles, DTRA for now would prefer to harvest existing data. “In any other city exposed to radiation, there would have been cleanup efforts that disturbed the exposure patterns, but that's not the case with Pripyat,” Pace says. DTRA has asked scientists who work in Pripyat to collect samples and report the results to the agency. Initial studies will not involve tracer gases. “Nor would we intentionally release radioactive materials,” says Pace.

A measure of good may yet come out of Pripyat's eldritch fate. “Pripyat is not a mockup. It is not a sterile façade of buildings erected for the purpose of blasting particles through its empty spaces,” says Chesser. “Bicycles, pianos, libraries, and baby dolls decaying through 19 winters are there to remind us that learning from this event really matters.” Pripyat would be a good laboratory, he says, precisely because it is real.

13. BRUCE ALBERTS INTERVIEW

# Attention, Class: A Departing NAS President Speaks His Mind

1. Jeffrey Mervis

Bruce Alberts may be stepping down as president of the U.S. National Academy of Sciences. But he's a long way from retiring

Bruce Alberts came to the National Academy of Sciences (NAS) hell-bent on improving U.S. science and math education. Twelve years later, as he wraps up his second term as the academy's 20th president, Alberts admits that the country's educational system is still broken. But he hasn't stopped trying to fix it.

Along the way, he's also strengthened the academy's position as a respected, independent source of advice to the U.S. government by reducing the turnaround time on many of the 200-odd reports churned out every year by the National Research Council (NRC) that he heads. Says presidential science adviser John Marburger about NRC's 2001 report on climate change, “It gave them credibility with the Bush Administration and increased their inclination to use the academy more often.” Former Clinton science adviser Neal Lane says the country “is indebted to Bruce … for his dogged determination to improve American science and math education, and for his commitment to international cooperation in science.”

On 1 July, Alberts will return to his beloved University of California, San Francisco (UCSF), where he'll reclaim his old job as professor of biochemistry and biophysics, sans department chair. And in case anyone thinks that the 67-year-old biochemist has lost any of the spark that brought him to the nation's capital in 1993, his official portrait unveiled last week (below) should put such notions to rest. Its most prominent feature is a tie festooned with bright yellow pie faces with protruding tongues that depict a range of moods. The neckwear pokes fun at the people in this town who take themselves far too seriously. It's also a sign that Alberts is leaving NAS older and wiser—but with his spirit intact.

On 6 May, on the heels of his final annual NAS meeting, Alberts sat down with Science to discuss his accomplishments and failures as head of the self-elected meritocracy that stands as the country's most prestigious scientific organization. He spoke of the threat to science from advocates of intelligent design, of the need to better manage the U.S. scientific enterprise, and of the prospects for China and India becoming the next great scientific superpowers. Here are excerpts from that interview with Deputy News Editor Jeffrey Mervis.

• On the debate over teaching evolution:” It says we've failed as scientists and science educators to convey the nature of science and its values to the American public, despite our world leadership in science and technology. … We've got to pay more attention to the education of young people and completely transform the way we teach introductory science at the college level. We are failing to make people understand what science is, or why they should care about ⁁]it. … We all fear that this movement toward a biblical interpretation of scientific facts will eventually make us look like some of the countries in the Middle East. If we're going to remain a world leader, we're going to need all the scientific rationality that we can muster.”

• On why education reform is so difficult: “We all think we understand education because we did well ourselves. It worked for us, and we think it should work for everybody else. But that's a big mistake. Half the brilliant students who come to Harvard planning to major in science drop out in the first year or two, because they don't get real science in their intro courses. Instead, they get huge amounts of knowledge that they must memorize before they can get to the good stuff, the hands-on and interactive courses.

We know what to do, and many of the small liberal arts colleges are doing it. But many of the large universities, with some notable exceptions, are not taking it seriously. … The incentives are wrong. Someone has to tell the department chairs that getting the resources they want—for equipment, graduate students, and so on—is going to depend on how they teach undergraduates. If you take away the money, the faculty will respond. I've learned that from spending 30 years in academia.”

• On advising the government: “The Bush Administration [in 2001] asked us 14 specific questions about climate change, and I give them credit for asking. They didn't have to. … There are other problems that have arisen, and we're trying to help with them. For example, people keep saying that climate change isn't real, and that the science isn't there. We've answered that question, and we're going to continue to insist on those answers, whether they like it or not. … There are many things we'd like to do that we haven't been able to. We'd like to do a major study on nuclear power—the safety issues and where we as a country should go. But none of us have been successful, over four administrations [two for Clinton, two for Bush], in getting anybody to ask us to do that. And I don't know why they're not interested. … It's obvious that the Department of Energy has to ask us to do it. Otherwise, it doesn't make any sense because they won't listen to what we've come up with.

One big mistake I made as NAS president was to hold a competition within the academy for topics that we should study. We came up with lots of good ideas. But there was no client for them, so they had little effect. Getting an agency to put up even a little money for a study makes a big difference in their interest.”

• On recreating an Office of Technology Assessment (OTA): “After Congress abolished OTA, we became the only show in town. We didn't like it, and we've tried to fill the gap, but we can't do everything. [At the same time], the idea of recreating it doesn't seem to have any political capital around here. We're not opposed to it, but you want to fight the battles that you think you can win.”

• On open access to journals: “I think that the community should push for access to scientific information as quickly as possible. We tried [with the Proceedings of the National Academy of Sciences] to see how short we could make it. We actually tried only a 2-month delay. But the next year a number of librarians told us that they would wait the 2 months and not subscribe, saving the money for other journals. And so with regrets, our publication committee decided to let it slip to 6 months. It's an experiment, and maybe someday we'll move it ahead to 5 months. But 6 months has allowed us to maintain our subscription base. In fact, for 146 countries it's free immediately. But for scientists in the countries that can afford it—U.S. and Europe and Japan—we ask them to pay.”

• On changing the way the academy does business: “We've tried to experiment, including some studies where the committees didn't even meet. But it doesn't work. The kind of thing we do needs that personal interaction. We get people together who don't know each other, and we create something different. For example, we did a report on the future of developmental toxicology, and we had scientists from both camps. … The first meeting was like Greeks talking to Romans. They didn't have a common language. It takes a couple of meetings, and some meals, before people get comfortable. And in the end they produced something unique. But you can't do that on the Internet.

There are a lot of tricks to the trade. A good chair knows how to call a coffee break when things aren't going well so that people can work out their differences. People want to see body language. … We've been pushing the envelope to do things faster, and we're going to keep trying.”

• On the impact of 9/11 on scientific openness: “I think it's been a disaster. We've hurt security by not giving visas to leading foreign scientists, insulting our friends, and sending their students to other countries. Our tremendous scientific vitality is based on mixing the best talent from around the world. Twenty-five percent of the NAS members were born in another country, and they are our best diplomats. We're jeopardizing that by creating barriers that make no sense, like requiring students to promise that they won't stay here. It should be the reverse.

We have this broken system, and after 9/11 we're enforcing these rules in the name of national security. But what we're doing is the opposite of national security. I can't imagine a more effective way of losing our scientific leadership than closing down this country to scientific exchange. … And if and when we do get the problem straightened out, all our university presidents will have to go to India and China and solicit students, and tell them that they are now welcome. That's crazy.”

• On the rise of science in Asia: “It seems likely to me that China or India will become the dominant scientific power. They take science and technology seriously, their young people are hungry to learn it, and they have such large numbers of people. But as we all know, there are many ways to make a mess of it. My favorite example is the recent science strike in France. They want more resources for science, which is good. But at the same time, you'd hope that they could adjust their system to make it more merit-based. Now, after your Ph.D., the first job gives you lifetime tenure. That's nuts. That's the perfect way not to run a scientific system. So I think the countries that will lead the world in science and technology are not just those with the most people. That's important. But you also need a system that allows the most talented people to have access to what they need to function effectively. Encourage the collision of ideas, and reward risk-taking and innovation. The United States is trying to do those things, too, but not well enough.”

• On his future: “My first year here was really hard. It was overwhelming. It was only after 4 years that I even started to think about staying. I hoped that in my second term I could do a lot, including fixing education. But I ended up spending most of my second term on international science. Now I hope to remedy that, starting in July, when I go back to UCSF. I'll be paid to focus on educational issues. One thing I want to do is stimulate better science by mixing people up, exposing them to new ideas, and helping them make new connections. As a young scientist, you have to be dragged out of your hole. But at the academy we've been doing that with our Frontiers of Science program and Keck Futures Initiative. I don't see why that sort of thing can't be done on the UCSF campus, or in the Bay Area.

I'm also trying to think of new models for scientists at the end of their careers. Continuing to run a lab and competing for grants until my third renewal is turned down and I have to leave in disgrace is not the way to go. We can't maintain an innovative system unless the old scientists become mentors and make way for the next generation. How do I get credit for this? I was president of the academy, so I don't need the credit. The worse way is to put your name on their paper. But why can't there be a second way, that also goes into the database, for people who really helped make things happen? I'd be proud if, after 10 years, you could find 30 papers that I had helped people to do good science.”