# News this Week

Science  02 May 2003:
Vol. 300, Issue 5620, pp. 714
1. INFECTIOUS DISEASES

# Battling SARS on the Frontlines

1. Dennis Normile

HONG KONG—This city, with its population of 6.7 million, has so far borne the brunt of the severe acute respiratory syndrome (SARS) epidemic. At the beginning of this week, it accounted for 1557 of the world's 5050 cases and 138 of 321 deaths. It has also become a laboratory for the epidemiology, etiology, and treatment of SARS and for the public health response to it. Hong Kong's biomedical researchers have been in the thick of the race to understand and treat the disease. And last weekend researchers from Hong Kong, along with colleagues from mainland China and Taiwan, gathered at a symposium here to review progress and assess future challenges.

The mood was cautiously optimistic. “We seem to be starting to bring the outbreak [in Hong Kong] under control,” says Yuen Kwok-yung, a microbiologist at the University of Hong Kong (HKU). For the past week, fewer than 20 new cases have been reported daily. But SARS remains a serious threat. Although the majority of those contracting the disease recover, researchers admit that they desperately need treatments. And many questions remain about the transmission and behavior of the virus. “It's still very early days,” says Malik Peiris, an HKU virologist who is part of a group that first identified the coronavirus responsible for the disease.

One encouraging sign is that the disease has not spread as far or as fast as some feared. Epidemiologist Thomas Tsang, a consultant to the Hong Kong Department of Health, notes that a 1968 outbreak of influenza that caused 700,000 or more deaths spread virtually throughout the world in just 6 to 8 weeks. By contrast, the first cases of SARS surfaced in China in November 2002; 5 months later the disease has spread to 26 countries, only seven of which have more than 10 cases each. “It does look like SARS is much less contagious than flu, given the higher volume of international travel these days,” Tsang says.

Scientists are puzzling out explanations for the notable surges of disease in Hong Kong, where two major clusters account for a large proportion of infections. The first occurred in early March when the illness was still a mystery. A patient admitted to Prince of Wales Hospital was given a nebulizer to help him breathe, standard treatment for severe pneumonia. But nebulizers atomize respiratory droplets—now believed to be the main source of transmission—enabling them to float much further than they would typically be transported by a cough or sneeze. A week later, more than 112 health care workers and patients came down with the disease. “Everybody in that ward was infected, even mechanics who came to do maintenance work,” says Joseph Sung, chair of the department of medicine and therapeutics at the Chinese University of Hong Kong (CUHK), who is also affiliated with the hospital. Lapses in hygienic practices may also have helped spread the disease to health care workers outside SARS wards, says W. H. Seto, a microbiologist and specialist in controlling hospital infections at HKU.

Next came the outbreak at the Amoy Gardens apartment complex, where 321 residents were stricken. A recent government report suggests that contaminated fecal matter leaking into a narrow, poorly ventilated light court provided a reservoir that spread the infection through one building; it was then communicated by human-to-human contact through the neighborhood.

Since the Amoy Gardens spike, the number of new cases has tapered off. Tsang says the pattern suggests that vigorous monitoring, tracking of contacts, and precautionary quarantine can control—but not likely stamp out—outbreaks. Several symposium participants mentioned recent work suggesting that some people harbor the virus while showing minimal symptoms. “This is not comforting, but it is not surprising,” says Peiris. “There is no respiratory virus we know of that does not have a disease spectrum from asymptomatic to mild to severe,” he says. Tsang adds that such individuals could be a continuing source of the disease, as well as transmission from other regions and even from the original, presumably animal, reservoir.

Still open is whether the new coronavirus acts alone. In Hong Kong, “this new coronavirus is invariably found in patients with SARS and is sufficient for producing the disease by itself,” Peiris says. But that might not be the whole story. Hung Tao, a virologist at the Chinese Center for Disease Control and Prevention in Beijing, found evidence of chlamydia infection in lung tissue recovered from seven patients who died of suspected SARS. “We also think the coronavirus is the main causative agent, but there seems to be evidence for co-infection with chlamydia,” he says. Other groups have posited the involvement of the human metapneumovirus and other agents.

“It could be that the coronavirus may by itself produce the disease but it may also open the door for other viruses, or nonviruses, to aggravate the disease,” Peiris says. Or, he adds, these agents could increase the coronavirus's transmissibility. A key research question, he says, is whether other viruses or bacteria are involved in certain individuals, dubbed “superspreaders,” who seem to infect whole clusters of associates.

In terms of possible treatments, the good news, says HKU's Yuen, is that for those who have been hospitalized in Hong Kong “there is a 93% chance of recovery.” Patients who succumb typically are elderly, suffer other ailments, or have waited too long to seek treatment. Even so, the treatment regimen in Hong Kong—centered on the use of ribavirin, an antiviral, and steroids, which modulate the body's immune response—remains controversial. Outlining a paper, in press at The Lancet, recounting their experience with more than 300 patients at Prince of Wales Hospital, CUHK's Sung says he and his colleagues now believe the disease progresses through three phases. During the first week, patients have high fevers but few other symptoms. From days 7 through 11, x-rays show increasing spread of pneumonia through the lungs. The third phase involves pulmonary destruction. Sung says this pattern has led them to prescribe high doses of ribavirin during the first week to suppress the amount of virus in the body. During the second phase, steroids are used to limit inflammatory damage to the lungs. If the patient proceeds to the third stage, “there is not much we can offer,” he says.

Sung, for one, thinks the regimen is providing some benefits. But in vitro tests once the coronavirus was isolated indicate that ribavirin is not effective against the virus. And some wonder if patients are simply getting better on their own. Ralf Altmeyer, a virologist at the Pasteur Institute in Paris, says, “Steroids might be useful to reduce lung damage due to the inflammatory response, and some patients seem to respond to ribavirin, but it's difficult to reach a conclusion about the efficacy [of the regimen] unless clinical trials are done.”

“We hope we can try different regimens and see which actually works better,” says Yuen, adding that at least one Hong Kong group has designed a trial. But if new cases continue to drop, the scientists may not get the chance to conduct it—at least not in Hong Kong. The epicenter of the disease seems to be shifting to mainland China, taking with it the attention of the world's infectious-disease community. While bemoaning the human toll of the disease, Hong Kong's researchers acknowledge that they have enjoyed the spotlight. They have produced a stream of papers detailing the epidemiology, microbiology, and infection control aspects of the Hong Kong experience. If not for the outbreak, “guys like me would never have been able to author a paper in The Lancet,” says HKU's Seto, who has a paper in press on the infections of hospital workers.

2. INFECTIOUS DISEASES

# Hungry for Details, Scientists Zoom In on SARS Genomes

1. Martin Enserink,
2. Gretchen Vogel*
1. With reporting by Dennis Normile in Hong Kong and Ding Yimin in Beijing.

Every week brings new questions about SARS—and sometimes, glimmers of answers. Last week, many of the questions centered on the genome of the new coronavirus believed to be the culprit. Since a Canadian research group first posted its entire sequence online on 13 April, genome information for nine other virus isolates has become available on the Web. Researchers around the world have devoured the genetic code like a pack of wolves, searching for clues to the virus's origins, behavior, and future.

After press time this week, Science was set to publish online a paper analyzing the genome from the BCCA Genome Sciences Centre in Vancouver, as well as one from the Centers for Disease Control and Prevention (CDC) in Atlanta (http://www.sciencemag.org/feature/data/sars/). But as these papers neared completion, researchers in other labs had their computers parse all the available genomes for themselves, and on the Internet, subtle variations among the 30,000-nucleotide genomes were the subject of intense speculation.

Now that sequencing technology has become cheap and widely available, almost every country affected by SARS is sequencing its own version, says coronavirologist Willy Spaan of Leiden University Medical Center in the Netherlands. Shortly after the U.S. and Canadian teams, two groups in Hong Kong and one in Singapore posted online the sequences of their SARS virus isolates. But the big surprise was the flood of data produced by the Beijing Genomics Institute (BGI), together with the Academy of Military Medical Sciences.

Well known for its role in sequencing the rice genome, BGI was eager to work on SARS from the start, says director Yang Huanming—but lacking the government's blessing, it had a hard time obtaining samples. The researchers finally got them on 15 April, and within days they had posted in GenBank partial sequences of five isolates—with minimal fanfare to avoid offending authorities. Recognition came only after the Chinese government's about-face and the sacking of top health officials; on 20 April, Chinese Vice President Hu Jintao visited BGI to praise its contribution. “Now we are all welcome to contribute to SARS research,” says deputy director Wang Jian.

View this table:

In the analyses set for publication this week, CDC and the Canadian group confirm earlier findings, based on the sequence of small snippets of the genome, that the new virus is distantly related to all known coronaviruses and deserves to be placed in a new group of its own (Science, 18 April, p. 413). Both groups also show the genome organization is similar to that of known coronaviruses, and both identify a number of putative viral genes. But neither of the groups had a coronavirologist in its midst, says Spaan, who believes a more expert analysis of the two genomes—which his laboratory is currently undertaking—may reveal intriguing additional details about the viral machinery.

Whereas the two North American isolates differ in only eight nucleotides, a broader comparison across the 10 genomes now available shows much more variation. Although some variations may be due to sequencing errors, the data do suggest that the virus is mutating quite rapidly, says coronavirologist Luís Enjuanes of the Autonomous University of Madrid; variations occur both in the enzymes that help it replicate and the proteins that sit on its outer surface. That could mean the virus will prove adept at eluding drugs and vaccines, Enjuanes warns.

A close reading of the variation hints at other fascinating trends, says Henry Niman, a Harvard surgery instructor who operates an increasingly popular mailing list about SARS science. Last week, Niman and others were speculating how differences of a single amino acid among viruses isolated from different patients might explain variations in pathogenicity and infectivity. Others are skeptical, however. To show such correlations, you need data from many more patients, or animal tests, says Spaan. “There's a lot of nonsense going around at the moment,” he says. To get a better handle on the problem, the World Health Organization (WHO) in Geneva has plans to set up a SARS genetic diversity databank, linked to a clinical databank, says WHO virologist Klaus Stöhr.

It could swell quickly, because new genomes are churned out almost daily. The Genome Institute of Singapore alone has four more ready but not yet posted on its Web site, says executive director Edison Liu. After all, in the SARS era, says Liu, “it's just a hop, skip, and jump” from a patient's bedside to another sequence.

3. INFECTIOUS DISEASES

# WHO Wants 21st-Century Reporting Regs

1. Martin Enserink

Chinese authorities' initial dissembling over SARS—which included hiding patients in military hospitals and ambulances—is a model of how a country should not deal with a potentially deadly epidemic. Now the World Health Organization (WHO) is hoping such misguided behavior will provide much-needed impetus to revise a 52-year-old international treaty aimed at controlling global health threats. The revision, which has been wending its way through the bureaucratic circuit for eight tortuous years, is on the agenda for the next World Health Assembly, which starts on 19 May. WHO hopes that SARS has created enough momentum to get revised—and tougher—guidelines in place by 2005.

Under the proposed new treaty, countries would have to notify WHO immediately of dangers emerging within their borders or confirm an outbreak if the agency hears about it from a third party. “I think there will be a roll this year” at the assembly, says David Heymann, WHO's executive director for communicable diseases. “Everybody now understands the importance.”

Outbreaks of SARS and other diseases can be bad for business. Since as early as the mid-19th century, “disease diplomacy” efforts have helped countries overcome their natural reticence to report them. In 1951, members of the newly minted WHO adopted the so-called International Sanitary Regulations, later renamed International Health Regulations, which require countries to report cases of three historically important diseases: plague, cholera, and yellow fever. They also list the specific measures that WHO can implement to contain an outbreak, such as requiring vaccinations.

Most public health experts agree that the regulations are hopelessly outdated. Previously unknown diseases pop up almost yearly, old diseases resurface, and trade and travel have shrunk the world and vastly increased the speed with which bugs can spread. When SARS erupted, for instance, China had no legal obligation to report it, and WHO had to haggle for information.

Under the new plan, governments would have to report any “public health emergency of international concern,” whether the agent is known or not. To define that term, WHO has tested an algorithm that includes factors such as a disease's seriousness, unexpectedness, and potential for international spread. Each country would have to have in place a minimal surveillance system and a “focal point” for communication with WHO. The agency would help investigate and, if necessary, notify other countries if an outbreak was confirmed; it would then recommend control measures.

Max Hardiman, WHO's coordinator of the revision process, insists that countries will benefit from these changes. For one, they'll get help with an outbreak when needed, which could stem health and, in some cases, economic losses. After the 1994 outbreak of plague in India, for instance, many countries imposed draconian and largely unnecessary travel and trade restrictions that cost the Indian economy an estimated $1.7 billion. If instead of holding back information, India had collaborated with WHO, such restrictions might have been prevented, Hardiman says. On the other hand, Toronto officials' outrage at WHO's advisory against travel to their city, lifted on Wednesday, shows how much opinions differ on what constitutes a prudent approach. Whether the world will accept a more prominent role for WHO remains to be seen. “If countries really want to resist, we don't have a lot of teeth,” Hardiman concedes. But he hopes that peer pressure will spawn compliance. “The basic enforcement mechanism,” says public health specialist Stephen Morse of Columbia University in New York City, “is ‘Do unto others what you would want them to do unto you.’” 4. INFECTIOUS DISEASES # Avian Flu Outbreak Sets Off Alarm Bells 1. Martin Enserink The economic ramifications seemed bad enough. An epidemic of avian influenza virus roaring through Dutch poultry farms for the past 2 months—more than 20 million chickens have already been slaughtered—is expected to cost hundreds of millions of dollars. But in recent weeks, public health officials already on edge about SARS have reported an unexpected human health toll as well, sparking worries that the disease may transform into a global flu pandemic. Fortunately, say flu experts, the virus, which belongs to a strain called H7N7, is not very efficient at person-to-person transmission. But there is a small chance that it could become more contagious if it evolves over time or recombines with a human flu strain. The consequences, however unlikely, could be catastrophic, says virologist Klaus Stöhr of the World Health Organization (WHO) in Geneva. Labs participating in WHO's Global Influenza Surveillance Network are already developing a vaccine against the virus. “We have to give this top priority,” says flu scientist Robert Webster of St. Jude Children's Research Hospital in Memphis, Tennessee. Since it was first detected on 28 February, the outbreak of the so-called highly pathogenic avian influenza has spread to more than 240 farms in central and southern regions of the Netherlands and to several more in neighboring Belgium. Avian influenza has loomed as a major human health threat since a strain called H5N1 sickened 18 people and killed six in Hong Kong in 1997. Earlier this year, the same strain killed a 33-year-old man, also from Hong Kong, and sickened his son; beyond that, there are no human cases known (Science, 7 March, p. 1504). H7N7 and its close relatives seemed less worrisome because they had caused very little human disease in past outbreaks, says Eric Claas, a virologist at Leiden University Medical Center in the Netherlands who helped identify the H5N1 strain. Indeed, in a 5 March letter to the Dutch Parliament, agriculture minister Cees Veerman called the human health risks “negligibly small.” He was wrong. More than 80 poultry workers and others have developed conjunctivitis, an eye infection caused by the virus. In at least three cases, there's strong evidence that the person directly exposed transmitted the virus to a family member. And on 17 April, a veterinarian who had visited a poultry farm died of pneumonia in a hospital; an autopsy showed massive amounts of avian flu virus in his lungs, and other pneumonia-causing agents were ruled out. The vet had not taken oseltamivir (better known as Tamiflu), an antiviral drug recommended for anyone potentially exposed to the virus, which was “probably a bit sloppy,” Claas concludes. Still, the severity of his disease is “very unexpected,” Claas says, as is the number of people with conjunctivitis. Experts do not know why the virus is more aggressive than it was previously known to be. To reduce the risk of the avian strain recombining with a human flu strain, which could happen in someone who is infected with both, authorities are urging poultry workers to get vaccinated against the currently circulating human strains. But recombination can also occur in pigs, notorious influenza mixing vessels that are thought to have spawned the pandemics of 1918, 1957, and 1968. Researchers have already found antibodies against the virus in Dutch pigs—“a big concern,” Webster says. Stöhr, who also coordinates the scientific battle against SARS (Science, 11 April, p. 224), says that Dutch and Belgian authorities “are doing the right thing” by stressing the importance of antivirals and flu shots. Compliance and rigorous surveillance are crucial, he adds. In WHO's worst-case scenario, the virus evolves to transmit efficiently among humans, infects 10% to 25% of the world's population in a matter of months, and, like H5N1, has a death rate of 30%. That would make SARS look benign. 5. PLANETARY EXPLORATION # NASA Bails Out of French-Led Mars Mission 1. Charlene Crabb* 1. Charlene Crabb is a writer in Paris. 2. With reporting by Andrew Lawler in Boston. PARIS—France has never lacked ambition when it comes to space, but its efforts to forge its own path to Mars may now have suffered a mortal blow. Following budget cuts and restructuring at CNES, France's space agency, NASA has pulled out of the$350 million NetLander mission, a French-led collaboration to study the weather and subsurface geology of Mars.

The timing of NASA's decision, communicated to CNES in a letter received last month, had planetary scientists wondering, at least initially, if the decision was influenced by differences between the United States and France over the war in Iraq. The political climate could have had “an indirect effect,” asserts NetLander's U.S. scientific coordinator Bruce Banerdt of NASA's Jet Propulsion Laboratory in Pasadena, California. “Probably no one will complain in Washington that a French-led mission got canned.” The mission's European scientific coordinator, Philippe Lognonné of the Institute for Physics of the Globe in Paris, suggested that the recent political events may have “tipped the balance. And our allies at NASA headquarters may have stopped pushing for an international Mars mission with France.”

But officials from the two space agencies say the decision simply reflects their tightened research budgets. “There was no politics involved in this at all; this was a purely pragmatic decision,” says Orlando Figueroa, director of NASA's Mars program office. “It's a coincidence,” says Richard Bonneville, assistant director of space research and exploration at CNES.

NetLander's fate now rests with the European Space Agency (ESA). Last week, CNES and ESA researchers met to discuss ESA taking control of the NetLander program and hitching the payload to a possible 2009 ESA Mars mission. The CNES board of directors was due to meet this week to decide if it will provide $88 million to continue developing NetLander until its launch. Instead of scratching the martian surface, NetLander aims to probe the Red Planet's deep interior and monitor its atmosphere. Eight countries, including the United States and Germany, signed up to help France perch four stationary landers on Mars, each packed with a seismometer, ground-penetrating radar, a multispectral camera, and other instruments. Three of the landers would drop around the Tharsis Ridge volcanic region, and the fourth would plop down 10,000 kilometers away in the southern hemisphere. Collectively, the four would help answer fundamental questions about the planet, including whether or not liquid water lies beneath the surface. The original plan was for NetLander to ride with a French-led Mars orbiter mission, due for launch in 2007. But last year, reeling from internal upheaval and a slashed budget, CNES scrapped the orbiter. The agency scrambled to keep NetLander alive by looking for another Mars-bound vehicle launching in 2009, possibly NASA's Mars Telecommunications Orbiter mission. But from last November, NetLander scientists and engineers heard rumblings that any rescue by NASA was in trouble. The NASA letter that reached CNES headquarters in Paris last month stated that the risks and additional costs of accommodating the landers on the 2009 mission were too great. The letter also explained that the NASA-funded part of the lander collaboration would end as well. The agency estimated that extending the effort from 2007 to 2009 would add$10 million to the original price tag of $30 million. ESA may put off a final decision on whether to rescue NetLander until next year, says Marcello Coradini, coordinator of solar system missions at ESA. Even if ESA doesn't take over, NetLander will have one lifeline remaining. The Russian Space Agency has offered to launch the four landers aboard a Soyuz rocket in 2009, but at a cost of up to$50 million, which may still be too much for the troubled CNES.

6. STEM CELLS

# Oocytes Spontaneously Generated

1. Gretchen Vogel

In a development that could alter the ethical landscape surrounding human embryonic stem (ES) cells, scientists have found that mouse ES cells can develop into oocytes in culture dishes. It is not yet clear whether the cells can be fertilized and develop into embryos. But if so, and if human ES cells turn out to have similar powers, such cells might allow researchers to get around some of the expense and ethical questions that arise from using donated eggs for therapeutic cloning experiments.

The procedure is deceptively simple, report Karin Hübner, Hans Schöler, and their colleagues in a paper published online this week by Science (www.sciencemag.org/cgi/content/abstract/1083452). The cells developed spontaneously in dense cultures of ES cells. The trick was identifying the cells that took on characteristics of oocytes, says Schöler, a developmental biologist at the University of Pennsylvania in Philadelphia. He and his colleagues developed a fluorescent marker for early-stage germ cells, cells that may turn into sperm or eggs. After 8 days, up to 40% of the cells glowed green. “I'm sure a lot of people will realize now that they had these early germ cells,” Schöler says. But without the marker, “you wouldn't have any idea what is happening.”

The researchers began to spot other similarities to germ cell development: Groupings of ES cells seemed to act like follicles, surrounding some of the apparent germ cells and even producing estradiol, a precursor of estrogen. And after 16 days in culture, some of the oocyte-like cells expressed proteins typical of meiosis, the specialized cell division that sperm and egg cells undergo. After 26 days, some of the cells broke away from their companion cells, much as normal eggs depart from follicles.

Perhaps most surprising, after about 40 days in culture, structures that looked like early embryos appeared. Most common were rounded groups of cells that expressed proteins similar to those produced by 16-cell embryos. In a few cases, the scientists found structures that resembled blastocyst-stage embryos, the stage at which ES cells are derived. Schöler says the clusters are likely parthenotes, embryos that develop from unfertilized eggs. Normal mouse oocytes are known to form parthenotes in culture in response to chemical signals or temperature changes, but despite multiple attempts to implant them in a womb, none has ever survived to birth.

Developmental biologists say the observations are a tantalizing glimpse of what might be possible in cell culture. The markers the team developed provide good evidence that some ES cells are taking on characteristics of oocytes, says Azim Surani of the Wellcome Trust Institute of Cancer and Developmental Biology in Cambridge, U.K. But he and Alan Spradling of the Carnegie Institution of Washington in Baltimore urge caution. Spradling suspects that the cells “are not exactly equivalent to normal oocytes,” which require complex and delicate interactions with surrounding cells to develop properly. To really prove that they are oocytes, Surani says, the team needs to show that the cells can be fertilized and produce viable offspring.

Such experiments will have significant ethical implications, says bioethicist Arthur Caplan of the University of Pennsylvania, who helped advise Schöler. “This paper is an ethical earthquake,” Caplan says. If ES cells could be a source of human oocytes, scientists might be able to use them for nuclear transfer experiments rather than eggs from human donors, which are in short supply. And if the cells prove to be functional enough for nuclear transfer but not for producing offspring, they might answer one of the main arguments against therapeutic cloning: that it creates embryos only to destroy them. But if the cells are capable of being fertilized and developing normally, Caplan predicts, they will only increase the concerns of people opposed to ES cell research.

7. NEUROSCIENCE

# Old Neurons Revisit Their Youth

1. Greg Miller

It could be bigger than Viagra, and it already has drug companies drooling: a treatment that turns back time on the aging brain and makes old neurons act young again. Such a potion is still a long way off, but a team of neuroscientists reports an advance in this direction on page 812. The researchers applied tiny amounts of the neurotransmitter γ-aminobutyric acid (GABA) to neurons in the brains of old monkeys. The treatment restored the neurons' ability to distinguish the orientation of lines and the direction of moving objects, renewing long-lost powers of discrimination.

“Anything that can show a reversal in the effects of aging is really exciting and potentially beneficial,” says Julie Mendelson, a neuroscientist at the University of Toronto. She and others are optimistic—to varying degrees—that this kind of research will eventually lead to drugs that improve the lives of older people. Many sensory problems suffered by the elderly stem not from deterioration of the eyes and ears, but from declines in the brain regions that process sensory information.

For most of a macaque monkey's life, neurons in the primary visual cortex (V1) respond selectively to lines oriented at a particular angle and to bars moving in a particular direction. But in 2000, Audie Leventhal of the University of Utah School of Medicine in Salt Lake City and co-workers reported that V1 neurons in very old monkeys lose much of this selectivity.

In the new study, Leventhal teamed up with four colleagues in China to try to reverse this age-related decline in monkeys who were 26 to 32 years old, equivalent to 78 to 96 human years, Leventhal says. As expected, many cells in older animals responded to a range of orientations and directions. But when the researchers spritzed on GABA—the brain's major inhibitory neurotransmitter—or a GABA-like drug called muscimol, the proportion of cells selectively excited by a particular orientation nearly doubled and the proportion of direction-selective cells roughly tripled, approximating the proportions found in monkeys just 7 to 9 years old. GABA and muscimol had no effect on the selectivity of V1 neurons in the younger monkeys.

Leventhal suggests that GABA-dependent neural communication declines with age and that this decline is to blame for old animals' indiscriminate neurons. Indeed, when the team blocked GABA in younger animals, their neurons lost their orientation and direction selectivity, in effect aging them 20 years in an instant.

No one knows why GABA communication might decline with age, but Ulf Eysel of Ruhr University Bochum in Germany says some evidence hints that the inhibitory cells that use GABA are particularly sensitive to disruptions of blood flow to the brain. Over time these effects may kill cells or reduce their ability to make and release GABA.

Leventhal has “really got something,” says Donald Caspary, a neuropharmacologist at Southern Illinois University Medical School in Springfield. His group has found a similar decline in GABA-supported networks in the aging auditory system. As in visual areas, these networks are important for helping the brain extract information from a noisy environment, Caspary says. Their deterioration could explain age-related trouble following a conversation at a loud party or navigating through traffic.

A drug that restores lost GABA function in the appropriate brain regions without serious side effects could be a blockbuster, and Leventhal and Caspary both say they've received quite a bit of interest from industry. Benzodiazepines—a class of tranquilizers that includes Valium—boost GABA function, but they've yet to be tested for reversing the effects of aging. The idea of a pill for restoring lost sensory function is no longer a pipe dream, says Caspary. “When we first said this in the lab 10 or 15 years ago, we all burst out laughing,” he says. “But now it doesn't seem so ridiculous.”

8. DEVELOPMENTAL BIOLOGY

# Purified Signaling Protein Stimulates Stem Cell Proliferation

1. Jennifer Couzin

After decades of painstaking and often fruitless effort, a team of scientists has purified a versatile protein that stars in multiple cell dramas, from embryonic development to cancer. The protein, called Wnt, turns out to have yet another role, another team has found. It coaxes blood stem cells to divide briskly, lending hope that Wnt may make it easier to experiment on stem cells and perhaps apply them in therapy.

Both papers, which appear online this week in Nature, are the culmination of many years' work and a collaboration between neighboring labs at Stanford University in California. One lab was struggling with a persistent problem in biology: Stem cell pioneer Irving Weissman, Tannishtha Reya, and their colleagues wanted to boost the number of stem cells in a petri dish without letting them diverge into different kinds of tissue. Producing lots of undifferentiated stem cells is crucial for studying and guiding their development.

Nearby, developmental biologist Roel Nusse and his lab members were wrestling with their own albatross, the signaling protein Wnt, which had resisted all attempts to purify it. Humans produce at least 19 Wnt proteins, which guide the positioning of body segments during development and control genes that otherwise trigger cancer. “Many postdocs and students have broken their backs on this project,” says Ken Cadigan of the University of Michigan, Ann Arbor, of Wnt purification efforts.

But Nusse refused to abandon hope of purifying Wnt, which clung to the cell wall and crumbled whenever scientists tried to extricate it. The “eureka” moment came when Nusse and postdoc Karl Willert determined that the protein was water-avoiding, or hydrophobic—a trait that its genetic sequence hadn't suggested. But if not embedded in Wnt's DNA, where was this hydrophobia coming from?

The answer emerged when the researchers weighed the protein and studied its metabolic activity. A lipid molecule latches on to Wnt before it's shuttled out of the cell, making it hydrophobic, they found. Removing the lipid blunted Wnt's abilities. That “explains a lot” about the trouble researchers have had purifying Wnt, says Hans Clevers, an immunologist at Utrecht University in the Netherlands. Nusse's team shifted its purification plan to one for proteins bound to lipids and successfully isolated a mouse Wnt protein.

Weissman's lab, meanwhile, was closely following Nusse's progress. A few years earlier, the two groups had found that partially purified Wnt helps stem cells divide. More recently, the researchers had discovered that when levels of a protein normally triggered by Wnt, β-Catenin, are increased in stem cells, at least 30% of them remained immature and did not differentiate. But it still wasn't clear whether Wnt was acting alone, and Weissman's group anxiously awaited the pure protein. With unadulterated Wnt in hand, Reya, now at Duke University in Durham, North Carolina, and a graduate student added the protein to mouse stem cells from bone marrow, which generate a range of blood and immune cells. Over 1 week, Wnt-treated stem cells produced at least six times more daughter cells than did controls.

Stem cells boosted by Wnt or β-Catenin performed nearly as well as naturally occurring ones, Weissman's group found. The researchers infused the cells into mice whose bone marrow had been wiped out with radiation. As few as 45 of the Wnt-treated cells helped rebuild the animals' immune systems.

Although the infused cells did not mimic normal processes exactly, researchers are enthusiastic. “This is one of the first times that you see amplification of stem cell populations, which is what everyone's been looking for,” says Leonard Zon, a geneticist at Children's Hospital Boston. Other proteins may work in a similar way: Guy Sauvageau of the University of Montreal in Canada has seen promise in a protein called HoxB4. “We're close to being able to tell clinical people that yes, they now have proteins that will allow the expansion” of stem cells, says Sauvageau. He and others say that having a readily replicating supply of stem cells may bring therapy one step closer.

9. ARCHAEOLOGY

# Tortoise Pace for the Evolution of Chinese Writing?

1. Andrew Lawler

Writing did not emerge in China until 2 millennia after it appeared in what is today southern Iraq, about 5200 years ago. But a team of Chinese and U.S. researchers now proposes that writing in China went through a long, slow evolution that stretches back an astonishing 8000 years. They argue that etched marks found on tortoise shells in a Neolithic grave in Henan Province are the earliest known precursors to what became the system of characters—and that they likely were used for shamanic purposes.

That claim, made in the most recent issue of Antiquity, has won media attention on the BBC and U.S. network news, but so far it has earned mainly skepticism from many Western scholars. “There's nothing new here,” grumbles Robert Murowchick, a Boston University archaeologist. He and others dismiss the notion that these simple geometric signs can be linked to early writing. But the research is sure to fire up a longstanding debate about how Chinese writing evolved and whether religious practices spurred its origin.

The team led by Xueqin Li, a senior archaeologist at the University of Science and Technology in Hefei, drew on artifacts from a site called Jiahu on the upper reaches of the Huai River between the Yellow and Yangzi Rivers. Discovered in 1962 and partially excavated in the 1980s, Jiahu has been radiocarbon-dated to the 7th to 6th millennia B.C. by three Chinese laboratories. Along with house foundations, cellars, pottery kilns, and a huge variety of animal bones, archaeologists investigated 349 graves containing everything from an ancient seven-hole flute made from the bone of a crane to turquoise ornaments.

In two dozen of these graves, excavators also found tortoise shells, which in one grave replaced the skull. Recently, archaeologists visiting the warehouse where these objects are stored took note of their markings. Incised clearly on 14 of the shells were nine signs, with an additional two on bone; most date from 6600 B.C. to 6200 B.C.

Many of the marks are simple combinations of lines, but the authors contend that some resemble early characters for “eye” and for several Chinese numerals. “We don't say they are language or words,” says co-author Garman Harbottle, a chemist at Brookhaven National Laboratory in Upton, New York. “We do say these are signs.” He sees the marks as the earliest evidence of “a long line of experimentation with sign use.”

Scholars have long debated whether marks made on pottery beginning about 4500 B.C. were simply local signs of ownership or clan or were precursors to later Chinese characters found at the end of the Shang dynasty around 1200 B.C. Li's team argues that those marks are indeed linked to characters and that the Jiahu evidence pushes back their evolution substantially. “When all the scholarly work is done, we will be able to trace the road map for the development of written language” in China, predicts Harbottle.

Nonsense, say others. “There certainly is a long evolution,” says Murowchick, “but there is no evidence that these are the key links—it's just too early to reach that conclusion.” William Boltz, a specialist in early Chinese writing at the University of Washington, Seattle, is also wary of tying the Jiahu etchings to the Shang breakthrough. “The time scale is just too great,” he says, and there's no clear evidence that the Jiahu people were related to those of the Shang dynasty thousands of years later.

Perhaps the most intriguing aspect of the find, however, is why rather than how Chinese writing developed. In early Mesopotamia about 3200 B.C., scribes began to use tablets and a complex symbol system solely for accounting purposes, such as recording the number of sheep or goats bought (Science, 29 June 2001, p. 2418). In contrast, the early Shang characters concern only the king's connection to ancestral and magical powers, similar to the earliest New World writing, which is linked to royalty and the sacred calendar (Science, 6 December 2002, p. 1872). Many of the Shang characters were found on tortoise shells just like those of Jiahu 4 millennia earlier, and there is evidence that tortoise shells were used as early as the Neolithic for divination. “This is another reason why we can't dismiss the possibility that Chinese writing arose in a context of divination,” says Boltz. The genesis is particularly hard to pin down because many scholars assume that there were earlier writings—about unknown subjects—on perishable material such as bamboo.

More answers may lie in Jiahu, where only a small fraction of the site has been dug. Says Harbottle: “We need more excavations, and we need to find more early sites.”

10. SCIENCE PHILANTHROPY

# Gates Grows UW's Genome Program

1. Robert F. Service

It never hurts to have the world's richest person living down the street. Last week, the Bill and Melinda Gates Foundation awarded a $70 million grant to the University of Washington (UW), Seattle, to promote genetics research. The lion's share of the money—$60 million—will help pay for a $150 million, 265,000-square-foot (24,600-square-meter) building for the departments of genome sciences and bioengineering. The remaining$10 million will support genomics projects aimed at curing diseases rampant in developing countries. The new building should enable the university to add 14 new faculty members and up to 400 graduate students and staff members, says geneticist Robert Waterston, head of UW's genome sciences department.

The donation is the latest bit of good fortune for genome research in Seattle. Two years ago, the National Institutes of Health awarded the university $30 million as part of two national centers of excellence in genomics. And in January, the university lured Waterston, a prominent leader of the Human Genome Project, away from Washington University in St. Louis, Missouri. Eric Green, the scientific director of the National Human Genome Research Institute in Bethesda, Maryland, says the grant signals that “the University of Washington has emerged as one of the premier genetics and genomics programs” in the world. 11. PLANETARY SCIENCE # The New Race to the Moon 1. Andrew Lawler* 1. With Pallava Bagla in New Delhi, Ding Yimin in Beijing, Dennis Normile in Tokyo, and Govert Schilling in Utrecht. After a long hiatus of lunar exploration, a host of countries—and companies—will soon begin launching a motley flotilla of spacecraft to Earth's nearest neighbor Once it was the most coveted body in the solar system, courted in the public spotlight by two superpowers demonstrating their superior technological prowess. That proved a brief flirtation. Now, nearly 3 decades later, the moon is the goal of a new generation of suitors in what space scientists hope will become a more lasting and rewarding relationship. The first to come calling will be a European spacecraft, set to go off this summer. The first of two Japanese missions will be launched in the summer of 2004. Later this decade, India and China hope to prove their mettle by sending their first homegrown probes beyond Earth orbit to circle the moon. A U.S. company is planning to launch a probe this fall, one of several commercial ventures on the drawing boards. Even NASA is talking about rejoining the party in years to come, possibly with a sample-return mission. Conspicuously absent from the new race to the moon, however, is the kind of international cooperation that characterizes the planned exploration of Mars. Observers say that's due largely to the technical and political motivations behind most of the planned missions, which leave little room for the international scientific community to team up on joint projects. “The danger is that a lot of money will go toward doing science that has already been done,” says Alan Binder, founder and director of the Lunar Research Institute in Tucson, Arizona, and principal investigator for the 1998 Lunar Prospector orbiter. “I'm not impressed with these missions.” Others say that informal contacts will prevent redundancy. “There is some competition,” acknowledges Hitoshi Mizutani, head of the planetary science division at the Institute of Space and Astronautical Science (ISAS) in Sagamihara, Japan. “But ultimately, cooperation will enhance our understanding of our common target.” Even if science is not the driver, however, most researchers are elated with the new interest in our closest neighbor in space. “This certainly could lead to a minirenaissance in lunar studies,” says Bradley Jolliff, a geologist at Washington University in St. Louis, Missouri (see sidebar). ## Icy allure The moon was a popular destination at the dawn of the space age. In 1959, the Soviet Union flew a spacecraft past the lunar surface and then crash-landed a probe. Soft landings by robots followed. A decade after the first encounters, humans arrived. By the time geologist Harrison Schmitt climbed back aboard the Apollo 17 landing module in 1972, six NASA expeditions had gathered 382 kilograms of soil and rock, conducted hundreds of experiments, and left behind dozens of instruments. Soviet robotic missions—orbiters and landers—shuttled to the moon for more than 15 years, culminating in the 1976 return of 170 grams of lunar soil. But the moon's allure soon faded, as long-duration orbiting laboratories and scientific probes to distant planets became the fashion. The next suitor—Japan's Hiten probe—didn't arrive until 1990, orbiting briefly before crashing on the lunar surface. And the probe was designed mainly to test spacecraft systems. Not until the innovative joint NASA-Department of Defense Clementine mission of 1994, however, did scientific interest in the moon begin to revive (Science, 16 December 1994, p. 1835). That interest was heightened by NASA's Lunar Prospector mission in 1998, which detected water ice near both poles mixed in with rocks and soil. “We realized the moon was a much more complicated place that we thought,” says Jolliff. The moon also began to cast its spell on engineers looking for ways to test innovative spacecraft designs beyond Earth orbit without going too far from home and especially on fledgling space-faring nations such as India and China. ## Techno test The first flight in the next era of lunar exploration will be this summer's European Space Agency's (ESA's) SMART-1, short for Small Missions for Advanced Research in Technology. Due for launch in late July or early August, the$100 million mission—developed and built in only 3 years—will map the geology and mineral composition of the moon in unprecedented detail and look for ice on the bottoms of permanently shadowed craters around the moon's south pole. Packed aboard the 1-cubic-meter spacecraft are three toaster-sized instruments: an optical color camera with a 40-meter resolution, the first infrared spectrometer to be flown to the moon, and an x-ray spectrometer.

The flight will test new technologies for missions to orbit the sun and Mercury early in the next decade. “SMART-1 is going to the moon only to take us farther,” says ESA's Director of Science David Southwood. Scientists must also be patient, as the probe will take 16 months to reach Earth's nearest neighbor. Rather than using conventional chemical fuel, SMART-1's engine will use xenon ions in an electrical field generated by solar power. The gentle acceleration will gradually boost the craft's elliptical orbit around Earth until it reaches a point where the moon's gravity takes over. “By the time SMART-1 enters an orbit around the moon, it will have traveled about 100 million kilometers,” says project manager Giuseppe Racca. That's 250 times farther than the flight plans of the Apollo astronauts, who made the 400,000-kilometer trip in 3 days.

Japan's two lunar missions are intended to tackle large technical challenges as much as to study the moon, and both have been repeatedly delayed. The country's space leaders “recognized not only the scientific merits of moon missions but the role they could play in developing space technology and in attracting the interest of young people,” says ISAS's Mizutani.

The first effort, to go off in August 2004, is a $100 million mission called Lunar-A. Once in lunar orbit, the spacecraft will fire two torpedo-shaped probes that will burrow some 2 meters below the surface—one on the near side and one on the far side. Each will contain seismometers to measure moonquakes and sensors to monitor heat flow. The data, say Japanese researchers, could yield important clues to the moon's internal structure. Meanwhile, the orbiter will gather high-resolution images and transmit the data back to Earth every 2 weeks. In 2005, Japan will launch the most expensive lunar mission since the Apollo program—the$300 million Selenological and Engineering Explorer (Selene). The National Space Development Agency, which will merge with ISAS later this year, will provide its powerful H-II rocket to loft the 1.6-ton spacecraft into space. Selene, named for the Greek moon goddess, will orbit the moon at an altitude of just 100 kilometers.

Selene will train a bevy of spectrometers, imagers, laser altimeters, radar sounders, and magnetometers on the moon to collect the most detailed information yet on its topography, the elemental and mineral content of its surface, magnetism, and gravity. The spacecraft will also release two smaller satellites. One will transmit data to Earth, while the other will send a radio signal to both Earth and the relay satellite. By comparing the arrival times of the radio signals, researchers will be able to determine the moon's wobble as it orbits Earth and gather data on the density of the lunar interior. All these data should help resolve questions about the moon's origin and evolution.

Other instruments will study the plasmasphere, the ring of charged particles that surrounds Earth. One year into the mission, the lunar orbiter will release a probe to test controlled landing technologies for use on more distant bodies.

## New entrants

ESA and Japan are not the only players who see the moon as an important scientific goal as well as a critical test bed for new technologies. Earlier this month, Indian scientists lent their support to the nation's first mission beyond Earth orbit. The proposed $100 million effort would put a 250-kilogram spacecraft into orbit 100 kilometers above the surface. Although the government has not signed off on funding the mission, Krishnaswamy Kasturirangan, chair of the Indian Space Research Organisation (ISRO), says, “it is not a question of whether we can afford [to go to the moon]. It's whether we can afford to ignore it.” The probe, which would be launched in 2007 or 2008, would carry x-ray and gamma ray spectrometers to examine the moon's composition, as well as cameras to create the first digital elevation map with a high spatial and altitude resolution of 5 meters. Scientists participating in a recent meeting in Bangalore said that the 20-month mission also could identify a suitable spot for a future robotic lander. The scientists heard a presentation by a 15-member lunar mission task force headed by George Joseph, former director of ISRO's Ahmedabad-based Space Application Centre, although the government has not yet released the panel's 80-page report. It is “certainly doable,” says M. G. K. Menon, a physicist and former science minister. “All the technological wherewithal already exists.” In language reminiscent of the Apollo era, Menon predicts that the mission “will excite the younger generation … while enormously increasing the confidence of the nation.” India's plans have already excited Canadian officials, who in March inked a cooperative agreement with ISRO. “It offers a wonderful opportunity—more so since these kinds of offers don't come often,” says Canadian Space Agency president Marc Garneau. His agency hopes to contribute a 10-kilogram payload costing less than$7 million, with the instruments to be chosen by the Canadian science community. Indian sources say that feelers have also been sent out to Israel and Germany, while ESA is interested in providing a far-infrared sensor for the mission.

China also does not want to miss the lunar boat. “We should catch the opportunity before the global program of returning to the moon is … in full swing,” says Luan Enjie, director of China's National Space Administration (CNSA). Like India, China plans to make use of its growing sophistication in launchers and satellites to secure a place in the big leagues of space exploration. This fall, for instance, it expects to put its first astronauts into Earth orbit.

China has pondered the possibility of launching a lunar probe since the 1990s. But the government finally appears willing to put some money behind the idea. Last month, CNSA announced that it had begun work on a three-part lunar campaign. The first phase sounds remarkably similar to India's plan: An orbiting spacecraft would obtain a three-dimensional map of the lunar surface and analyze soil composition and material distribution. Dubbed Chang'e 1, after the wayward goddess in Chinese legend who was trapped on the moon, Luan says this mission would be completed by 2010. Ouyang Ziyuan, chief scientist of the moon program, predicts that the mission could fly within 3 years of approval, barring technological problems. The probe would use the structure of the current Dongfanghong 3 satellite series and be launched aboard a Long March 3A rocket. The rest of the lunar campaign is sketchier, although officials say that it would likely involve roving robots followed by a sample-return mission.

Luan says that China already possesses the necessary technology to go to the moon. “The only thing that Chinese scientists need to do is to improve the current technologies to fit them into moon exploration,” he says. But whether China can produce cutting-edge science with such a mission remains a question. “It is not a problem sending something [to the moon] just as a symbol,” says Liang Sili, expert on rocket control systems with China Aerospace Science and Technology Corp. “But we need to explore the moon with some creative scientific goals. … We should try to do more things other than following others.”

One of those goals could be to exploit lunar resources. Luan cites the possible presence of helium-3, which could be used in future nuclear fusion reactors. The mineral and energy resources unique to the moon are important to supplement reserves on Earth, he says.

## Cash customers

Lunar resources also have caught the eye of U.S. entrepreneurs, who see moon missions as a profitable market. In November, a private company based in La Jolla, California, intends to launch a $20 million, 110-kilogram probe into lunar orbit carrying two cameras. One would provide 1-meter-resolution images—an order of magnitude sharper than currently available—while the other would be a video camera providing continuous movies of the lunar surface. TransOrbital president Dennis Laurie is betting that educators, advertisers, filmmakers, and video game developers would snap up high-quality lunar pictures. The orbiter also would release a small capsule that would crash-land with a payload containing everything from business cards to cremated remains. TransOrbital successfully conducted an Earth-orbit test of a probe mockup in December, and the November launch will use a modified Russian SS-18. “We're pretty much on schedule,” Laurie says, although he declines to discuss financing details. Others hope to follow. LunaCorp of Reston, Virginia, plans to put up a lunar mapping and telecommunications satellite that would weigh about 200 kilograms and cost between$20 million and $30 million. Launched in pieces aboard the space shuttle and assembled on the international space station, the spacecraft would avoid the difficulties of the first few minutes of launch in Earth's atmosphere. LunaCorp chief David Gump hopes NASA, other space agencies, and scientists might buy services from the satellite, dubbed Supersat. But the Columbia disaster makes it unclear when the shuttle will resume flying, he adds. Conspicuously absent from the ranks of lunar explorers is NASA. “NASA doesn't care about the moon one hoot,” complains the Lunar Research Institute's Binder, whose Lunar Prospector mission, he maintains, was funded by the space agency thanks to pressure from outside scientists. “They want to go to Mars and fear the public might instead get interested in the moon.” But proponents haven't given up hope: This spring NASA will hold a competition for a$650 million solar system exploration mission (see sidebar), and at least two teams will propose a lunar-sample return. NASA space science chief Ed Weiler says that lunar missions “will just have to compete like everyone else.”

Competition will also be part of the new moon race, although it may lack the drama of the first, Cold War-fueled contest. That's just fine for researchers, so long as the data are fresh. The moon has lost none of its charms over the years. If anything, to a new generation of admirers, absence has made the heart grow fonder.

12. PLANETARY SCIENCE

# Moon Maintains Its Mysteries

1. Andrew Lawler,
2. Govert Schilling

Telescopes have been trained on it since Galileo's day, and dozens of spacecraft have flown by it, orbited around it, and landed on it. And a dozen humans—including one geologist—have walked on the surface and brought back soil and rocks. So the moon must be the best understood of all places beyond Earth, right?

Wrong. “That's an illusion,” says Carlé Pieters, a planetary geologist at Brown University in Providence, Rhode Island. “We're doing lunar exploration backwards—we have a wonderful set of samples, but a poor foundation for a global assessment.” Despite a wealth of data, critical questions remain about the moon's formation, the massive early impacts on its surface, and its chemical composition. “Thirty-five years after Apollo, our knowledge about the moon is still surprisingly incomplete,” adds Bernard Foing, project scientist for the European Space Agency's SMART-1 mission to be launched this summer.

And yet, the moon matters. Understanding its geology is critical for understanding the formation of the terrestrial planets and the solar system—as well as the Earth-moon system. The great age of the moon's surface—the result of limited geological processes over the past few billion years—makes it an invaluable aid in reconstructing the period when planets were forming. In a sense, says Foing, “the moon is a book of history of the solar system.”

The Department of Defense's 1994 Clementine mission and NASA's Lunar Prospector, launched in early 1998, mapped numerous elements across the surface and apparently detected ice deposits in permanently shaded craters near the poles. They also mapped the moon's gravitational field. But more details are needed on the location and concentrations of minerals, as well as high-resolution topographical maps and high-spatial-resolution imagery to pick out details such as fissures and boulders down to 1 meter in size, says Pieters.

The planned European, Japanese, Indian, and Chinese missions will focus on these areas. In addition, Japan's Selene mission next summer will provide a more detailed gravity map, revealing telling details about the interior. But Lunar Prospector principal investigator Alan Binder says that some of these missions—such as SMART-1—will duplicate data and have elliptical orbits that will limit coverage.

One of the most intriguing spots on the moon is the South Pole-Aitken Basin, a vast crater that stretches from 2200 kilometers to 2500 kilometers in diameter and is 15 kilometers deep. Examined by Clementine and Lunar Prospector, it is likely the oldest, biggest, and deepest such basin in the solar system. The impact that created it was massive enough to expose material in the mantle, geologists believe. Such material could prove extremely valuable in understanding the lunar interior and its early development.

Dating the age of the basin is also high on the list of solar system scientists. Heavy bombardment in the inner solar system ceased 3.85 billion years ago. Only after that date could life have developed and persisted. Current estimates make the Aitken Basin about 3.9 billion years old. That date, if correct, would lend credibility to the theory of a fast and furious period of bombardment late in the formation of planets. Alternatively, if the crater is much older, it points to evolution rather than cataclysm. “But you can't get the age without a sample,” says Pieters.

To that end, a U.S. National Research Council committee set up to prioritize solar system exploration last year listed a basin-sample-return mission as its second highest priority for future mid-sized missions, behind a flight to Pluto and the Kuiper Belt (Science, 19 July 2002, p. 317). Lunar advocates are hoping that the report will give the moon an edge in an upcoming NASA competition for a new solar system mission.

13. BIOTECHNOLOGY

# Hatching the Golden Egg: A New Way to Make Drugs

1. Joe Alper*
1. Joe Alper is a writer in Louisville, Colorado.

After 2 decades of work, researchers have succeeded in creating gene-altered chickens that can lay eggs containing human proteins

Chicken farms are among the most pungent places on Earth, but if a few hardy entrepreneurs have their way, it may soon be the sweet smell of success that wafts from buildings housing thousands of transgenic hens. As many as a half-dozen small companies are hoping to turn the common chicken into a pharmaceutical bioreactor, one that can meet the growing demand for protein-based human therapeutics.

The idea is deceptively simple: Insert human genes into chickens and get them to make human proteins in egg whites. “Transgenic chickens should be near-perfect bioreactors for making large amounts of pure recombinant proteins,” says Ann Gibbins, an avian biologist at the University of Guelph in Ontario, who did the pioneering studies in the field and trained many of those scientists now achieving success. But, as she found out the hard way, putting the scheme into practice has been tough. Now, after 2 decades of struggles, researchers are reporting numerous successes in making transgenic chickens. The next-generation bioreactor should follow close behind.

As far as the drug industry is concerned, such transgenic chickens will most certainly be laying golden eggs. It's just a matter of time, drug companies say, before they will be riding a wave of new health products based on manufactured human proteins. Already, they have concocted more than 300 human antibodies in the lab and begun testing them against a variety of human ailments—from cancer to viral infections to immune system disorders. But the companies need a better way to make these products. Current methods can be inefficient: The biotech firm Amgen, for example, has been having trouble meeting demand for its arthritis medication Enbrel, which contains a human protein made by Chinese hamster ovary cells.

The math of making drugs in chicken eggs is appealing. Each commercial hen lays about 250-plus eggs a year, at a nickel apiece. Each egg contains nearly 4 grams of egg white, comprising a mere eight proteins. Get a transgenic chicken to add 100 milligrams of a recombinant protein to that mix, Gibbins says, and the final cost for purified protein should be about \$10 per gram, 100-fold less than the cost of current systems using cultured mammalian cells. Compared with other proposed animal production systems—cows or goats, for example—chicken flocks are easy to ramp up in months. There are other benefits as well: Commercial egg farms are already secure enough that a transgenic chicken, for instance, is unlikely to escape into the wild and breed. And there's a bureaucratic comfort factor: The U.S. Department of Agriculture and the Food and Drug Administration are familiar with eggs as bioreactors because many vaccines, including those against influenza, are already produced in chicken eggs, although not transgenic ones.

Human biology also suggests why egg-reactors should work better than alternatives such as bacteria or nonhuman mammalian cells. “Chickens add sugars to finished proteins in much the same way that humans do, something that can't be said for most other protein production systems, current or envisioned,” explains virologist Bill MacArthur, president of GeneWorks, a chicken research firm in Ann Arbor, Michigan. Proteins finished with the human glycosylation pattern are more active biologically.

But for all the potential benefits, “creating a transgenic chicken has proven to be a far greater challenge than we ever expected,” says Gibbins. Although reports of success have appeared in the literature every couple of years, none has panned out. Either the introduced genes failed to carry over into later generations or the chickens failed to produce the desired protein in their eggs. “It left us all thinking that it was going to take some kind of magic to ever make a transgenic chicken egg.”

The magicians have arrived. At least three research teams using different methods have now shown that they can make transgenic chickens in proof-of-principle experiments. “Nobody has a transgenic chicken ready to produce a pharmaceutical today, but the field has now gotten to a place where this will happen,” says developmental biologist Jim Petitte, whose group at North Carolina State University, Raleigh, developed a transgenic chicken that produces the bacterial enzyme β-galactosidase in its eggs.

## Needle in a haystack

Much of the trouble in creating a transgenic bird arises from some unusual features of avian reproductive biology: the sheer size of the egg and the difficulty of harvesting one before it has begun growing into a chick. The most common way to make a transgenic animal is to harvest a newly fertilized egg and inject foreign DNA directly into this single cell's nucleus using a microscope and an ultrathin syringe needle. Eggs are easy to harvest when laid, but by that stage the zygote has already grown into a 60,000-cell mass. Harvesting the zygote from the chicken's oviduct, difficult and expensive in itself, does yield a single cell, but one almost impossible to find within the viscous, yellow yolk.

Helen Sang, an avian molecular biologist at the Roslin Institute in Midlothian, U.K., is one of the few researchers in the world to have figured out how to harvest a newly fertilized zygote and inject it with DNA or manipulate the 60,000-cell zygote of a freshly laid chicken egg and still get it to hatch into a transgenic chicken. Working with researchers from Viragen in Plantation, Florida, Sang has successfully created transgenic chickens that express a green fluorescent protein and are able to pass the introduced genes on to subsequent generations.

In this proof-of-principle study, Sang's team used a lentivirus-based gene delivery system engineered at Oxford BioMedica in London to transfer the gene for green fluorescent protein or for β-galactosidase into the DNA of chick embryos from freshly laid eggs. The injected embryos were then transferred to a host shell whose top had been carefully removed. Plastic wrap sealed the breech in the eggshell, allowing about 30% of the reconstructed eggs to develop and hatch normally. Of the eggs that hatched, 10 were transgenic roosters, evidence that the introduced gene had incorporated itself into the chicken's DNA. Upon mating, each rooster successfully produced up to 29% transgenic offspring expressing the green fluorescent protein or β-galactosidase, which Sang characterizes as “remarkably successful … far better than anything we've seen before.” Using Oxford BioMedica's lentivirus system, she is now developing a transgenic chicken designed to express a “clinically relevant human monoclonal antibody.” By linking the gene for this antibody with the promoter for one of the egg-white proteins, Sang hopes to have it expressed in commercial quantities within the egg white.

Rather than try to manipulate a chicken egg, researchers at BioAgri in City of Industry, California, are working with rooster sperm. The key to this strategy is a monoclonal antibody, developed by BioAgri scientists, that binds specifically to the surface of sperm and allows DNA linked to the antibody to enter the sperm cell and incorporate itself into the sperm's genome.

Using this sperm gene-transfer technique, developmental biologist Jin Qian and his colleagues have created two different transgenic chickens that produce human interferon α and interferon β. The BioAgri researchers have shown that the interferon genes are stably transmitted across two generations of offspring so far. In addition, their subsequent study using green fluorescent protein linked to the lysozyme gene promoter demonstrated successful gene expression in egg white. The chicks from that experiment are nearing maturity and will be bred to see if this gene is passed to the offspring.

Based on their successes so far, both Viragen, which supports Sang's work, and BioAgri are now negotiating with pharmaceutical companies to develop transgenic chickens that will make human proteins for clinical development. Petitte, who is not associated with either firm, wouldn't be surprised if these deals were inked soon, a sign that this field is finally realizing its promise. “In the late '80s, lots of pharmaceutical and biotech companies had transgenic chicken programs because everyone saw the potential for lowering production costs, but they all got off the field when their programs went nowhere,” says Petitte. “Fortunately, some of us were stubborn, and now the payoff is here. It's time to be optimistic.”

14. COSMOLOGY

# With Its Ingredients MAPped, Universe's Recipe Beckons

1. Charles Seife

Now that the Microwave Anisotropy Probe has nailed down what everything is made of, researchers are scrambling to figure out how it all came together

DAVIS, CALIFORNIA—A month after the first pictures from the Wilkinson Microwave Anisotropy Probe (the satellite formerly known as MAP) pinned down the fundamental constants that describe the cosmos, prominent stargazers and theorists gathered here to gear up for the next phase in cosmology.* All agreed that their field was far from played out.

For years, physicists using balloons and ground-based observatories have probed the cosmic microwave background (CMB), the glow left over from 400,000 years after the big bang, for clues to the “shape” and composition of the cosmos. When the MAP data came out in February (Science, 14 February, p. 991), that era in cosmology ended. Scientists now know that the universe is geometrically “flat,” that it is made up of about 27% matter and 73% dark energy, and that it is 13.7 billion years old. They have figured out, with great precision, how fast it is expanding. Now, they say, it's time to move on.

“I cannot overemphasize how important the MAP data is,” says Max Tegmark, a cosmologist at the University of Pennsylvania in Philadelphia. “But with MAP, it's not like it's all over and we should all switch into biophysics.” On the contrary, he argues, “cosmology is becoming more and more fun.” The pursuit of fun is taking scientists in two directions: forward in time from the CMB to glimpse surprising insights about when and how the first stars and galaxies were born; and backward, to give inflation—a class of theories that describe how the universe expanded less than 10−32 seconds after the big bang—a long-awaited reality check. On both fronts, new observations of cosmic phenomena are beginning to take researchers into unexplored territory that the CMB alone could not reveal.

## Premature star birth

To illuminate this terra incognita, astronomers start with light. MAP's exquisite pictures of the CMB were photons bouncing off a cloud of hot gas, mostly hydrogen, that filled the universe when it was about 400,000 years old. During its 14-billion-year journey, stretched and cooled by the expanding fabric of the universe, the light traveled through other clouds of hydrogen. Early on, those clouds consisted largely of electrically neutral intact atoms that had condensed from their component protons and electrons as the universe cooled. Later, energy from stars, galaxies, and other energetic cosmic objects stripped the electrons from the hydrogen nuclei, a process known as “reionization.” Because neutral hydrogen is opaque to some wavelengths of light, determining how long neutral hydrogen “fog” suffused the universe before reionization burned it away should tell scientists when stars and galaxies ignited—a landmark in cosmic history.

By studying variations in the CMB's temperature and polarization, MAP astronomers figured out how much of the CMB had been absorbed by the neutral hydrogen fog during its long journey. To their surprise, their calculations showed that the fog began to burn off when the universe was a mere 200 million years old and then lifted rapidly. That implies that stars and galaxies must have been forming hundreds of millions of years earlier than most astronomers thought. “Literally, before 1 month ago, everyone agreed that reionization occurred at about [2 billion years after the big bang],” says Joe Silk, a cosmologist at the University of Oxford, U.K. “Something has to have been going on in protogalaxies.”

To burn off fog so efficiently, says Columbia University physicist Zoltán Haiman, star formation in the early universe must have been radically different from what it is today. Martin Rees, Britain's Astronomer Royal, thinks that the first stars may have emitted much more hydrogen-ionizing ultraviolet (UV) radiation than present-day stars do. “Perhaps earlier stars were much more massive,” he says. “They could have been 100 solar masses or more, efficiently producing UV—very hot.”

Lam Hui of the Fermi National Accelerator Laboratory (Fermilab) in Batavia, Illinois, thinks the generation of giants may actually have been two generations. By studying light from quasars, energy-spewing galaxies that shine like beacons through intergalactic gas clouds, Hui concluded that the gas clouds are too warm to have been reionized in a single burst. He thinks there might well have been an early reionization, a cooling when the fog resettled, and then a re-reionization hundreds of millions of years later as the ancient generation of stars burned out and the newer ones began to blaze.

In turning to quasars, Hui is part of a trend. MAP data reveal only the general level of fogginess throughout the universe's history. To pinpoint how and when the first stars lit up, researchers need to know when the pockets of fog were thickest and how the fog burned off. Because quasar light is more recent than CMB light and yields information about structures that are younger and smaller than the clumps in the CMB, quasars can give astrophysicists much better information about the clouds: their shape, location, age, and composition, as well as how clumpy they are and how fast they're moving. In the past few years, ground-based telescopes such as the Keck in Hawaii have yielded thousands of quasars for astrophysicists to analyze. The Sloan Digital Sky Survey, a massive 5-year effort to map about one-eighth of the sky (Science, 25 May 2001, p. 1472) is beginning to supply many more.

## A check on inflation

Those quasars may hold the key not only to the birth of stars and galaxies, but to the dawn of the universe itself. With their help, researchers hope to strengthen a bulwark of modern cosmology, the family of theories known as inflation. Inflation posits that the tiny universe expanded extremely rapidly for a tiny fraction of a second after the big bang. Among other things, that growth spurt explains the “flat” Euclidean geometry that MAP confirmed.

Although the inflationary theories all paint the same general picture, the hundreds of rival models disagree about the precise mechanism that set the inflation in motion and how it dissipated over time. So far, nobody has had enough data to weed out the unfit. But new observations are starting to change that, Tegmark says. “People are seriously worrying whether inflation theories are making sense or not.”

The worries center on a property called scale invariance. Imagine a pebbled beach where for every boulder one meter across, there are 100 pebbles a centimeter across, a million pebbles one micrometer across, and so forth. If you took a picture of such a beach, it would look much the same as a snapshot that an ant-sized being would take on the same beach. The distribution of rocks is said to be scale invariant.

When physicists looked at pre-MAP images of the CMB, that is what they saw. The biggest clumps of matter in the universe, enormous concentrations and rarefactions of plasma in the primeval cosmos, appeared to be scale invariant, at least within the error bars on the data. In mathematical language, the fluctuations in the CMB had a “spectral index” approximately equal to one.

Inflation, however, dictates that clumps of matter in the universe are almost, but not exactly, scale invariant. Now that MAP has shrunk the error bars, some analysts think they have glimpsed that “not exactly.” A clear sighting would mark a great triumph for inflation and would enable physicists to discard inflationary theories that predict different deviations.

Hoping to resolve the issue, researchers are once again turning to quasars. So far, the results have been inconclusive: Two groups analyzing the same quasar data have come up with starkly different answers. One, represented by Fermilab's Hui, sees no deviation from scale invariance. The other team, which included Princeton University's Uros Seljak, claims to have spotted not only a significant deviation from scale invariance but also a change in the spectral index over different scales, a quantity known as the “running” of the spectral index, far larger than most inflation models predict. If Seljak's team is correct, almost all inflationary theories can be ruled out right away. Most physicists, however, are skeptical. Hui suggests that the differences between the two groups' conclusions arise from differing assumptions about the properties of the telescopes as well as assumptions that went into the computer models that contribute to the analysis. “We're trying to get to the bottom of it,” he says.

Tegmark says the disagreement is a sign that cosmological debate has moved to a new level. He calls it “gastrophysics”: a stomach-churning realm in which tiny errors that were once too small to matter suddenly gain the power to make or break a theory. As measurements get better and more data pour in, physicists will bring those errors under control and chart exciting new territory. But for many, the wait is a strain.

In the electrified post-MAP atmosphere at Davis, patience was harder than ever to come by. In one session, discussing tantalizing CMB data hinting that the universe might have an exotic closed geometry like a doughnut, Princeton's David Spergel urged his audience to resist the temptation to overinterpret. “Don't obsess on this. Don't obsess on this. Don't obsess on this,” he warned. It remains to be seen whether anyone was paying attention.

• *The Davis Meeting on Cosmic Inflation, 22 to 25 March.