# News this Week

Science  07 Mar 2003:
Vol. 299, Issue 5612, pp. 1494
1. CLIMATE RESEARCH

# Right Direction, But a Long Way to Go for Bush's Plan

1. Richard A. Kerr

In a sympathetic but often critical review of the Administration's draft plan for climate research, a committee of the National Research Council (NRC) last week called for substantial revisions before the final plan emerges in just 2 months. Calling the draft “a solid foundation” for the $1.75 billion Climate Change Science Program (CCSP) instituted by President George W. Bush last February (Science, 22 February 2002, p. 1439), the committee in its next breath declared that the “draft plan lacks most of the basic elements of a strategic plan … neither an explicitly stated vision nor a coherent set of goals are consistently presented.” “It's a fair challenge,” says CCSP's director, James Mahoney, assistant secretary of commerce for oceans and atmosphere. He promises that “the final plan [will] be much different than the draft plan,” although he provides no particulars of how it will reflect the criticism. In addition to more vision and focus, the review calls for mustering more resources—fiscal, human, and infrastructural—for the president's broadened attack on the climate problem. It also calls for a better plan to coordinate the 13 federal agencies tasked with producing climate science that is more relevant to policy. CCSP is George W. Bush's modest enhancement and substantial refocusing of the Global Change Research Program (GCRP) that his father began in 1989. So far,$20 billion has been spent under GCRP. The president's creation last year of a $40-million-a-year add-on, the Climate Change Research Initiative (CCRI), was intended to focus on finding science-driven answers for policy problems such as water resource management and coastal planning. These answers would be of more immediate value to decision-makers than GCRP's curiosity-driven research and would be produced quickly, within the next 2 to 4 years. As drafted, the CCSP plan lacks the needed focus to either size up the problem or inform policymaking, according to the NRC review. The draft plan places a commendable emphasis on producing scientific results useful for decision-makers, says NRC committee chair Thomas Graedel of Yale University, and many key elements are scattered through the plan. But “the plan as drafted tends to lack a consistent guiding framework, a specific and ambitious vision, or a set of goals,” he says. For example, it provides no means for determining how something qualifies for the short-term CCRI versus the long-term GCRP. The NRC panel also found that the overall effort lacks sufficient funding. “This is a big plan,” says Graedel, “bigger than the one in place before.” It goes beyond study of the climate system, the focus of the old GCRP, to include societal and ecosystem impacts and an upgraded global climate observing system, among other enhancements. “It's clear to the committee,” says Graedel, that “if we expand the scale of the program, it's going to require expanding the resources available.” That isn't happening, the report notes. The committee found that CCRI's$40 million fiscal year 2003 request had ballooned to $182 million in the president's 2004 request. But the money came out of GCRP, the committee says; the total CCSP request rose only$2 million to $1749 million in 2004. Mahoney concedes that funding will be tight. “If you're going to change the direction of a program,” he says, “we have to be open to new things. We've got to adjust and prioritize” within the limits of available funding. Such priority-setting will require tight management and coordination. But the NRC committee wasn't convinced that CCSP's 13 agencies—funded through a half-dozen congressional committees—will be working closely toward the same goals. “It's not clear from the draft exactly where authoritative responsibility for the program will reside,” says Graedel. “This is a new start. It has an opportunity to redefine how agency operations are better integrated.” The panel's advice for coordinating priorities within CCSP is the establishment of “a standing advisory body charged with independent oversight of the entire [CCSP] program.” It wouldn't be the first attempt to coordinate interagency climate research programs. In the early 1990s, D. Allan Bromley, then director of the Office of Science and Technology Policy, revitalized the existing Federal Coordinating Council for Science, Engineering, and Technology (FCCSET, pronounced “fix-it”) to orchestrate large, multiagency programs, including GCRP (Science, 15 February 1991, p. 737). That has not managed to elevate the coordinated needs of the program over the agencies' priorities, according to many observers (Science, 13 July 2001, p. 199). “These are very difficult issues,” agrees Mahoney. Without promising independent oversight, Mahoney points to the management structure already in place that can move oversight questions as high as the Cabinet level. 2. HIV/AIDS # Vaccine Results Lose Significance Under Scrutiny 1. Jon Cohen Everybody expected that the results of the first test of an AIDS vaccine would be controversial. Many researchers, after all, had been skeptical about the value of the vaccine for years and were waiting to pounce. But few would have predicted the confusion that ensued when VaxGen of Brisbane, California, released its long-awaited findings on 24 February (Science, 28 February, p. 1290). Over the next few days, VaxGen officials contradicted each other about how the data were analyzed and reported and whether some of the conclusions are statistically significant. VaxGen's 5000-person study conclusively showed that the product failed to prevent infection with HIV, a result that sparked no disagreement. But company scientists claimed that analyses of subgroups revealed a statistically significant efficacy rate of 66.8% in blacks, Asians, and people of mixed race, and 78.3% in blacks alone. A few sharp-eyed scientists quickly questioned whether the company had calculated the statistics accurately. Bette Korber, an HIV geneticist at Los Alamos National Laboratory in New Mexico, and biostatistician Steven Self of the University of Washington, Seattle, independently concluded that the subgroup analyses appeared to be flawed. Each time researchers conduct a substudy of a data set, they increase the likelihood that an apparently significant result will be due to chance. To correct for this, statisticians assess a statistical penalty for each additional subgroup analysis. Specifically, statisticians adjust the confidence interval, which says with a given degree of certainty, or “P value,” that the result is not due to chance. Both Korber and Self concluded that VaxGen had not adjusted for the subgroup analyses in the results the company reported. On 26 February, Marc Gurwith, an infectious-disease specialist who heads the company's clinical trials, conceded to Science that because of a misunderstanding between the company's statistician and other scientists there, “the P values that were in the [24 February] press release were not adjusted.” Gurwith acknowledged that if a conservative correction were made, the significance of the result in blacks would disappear. Gurwith emphasizes, however, that statisticians do not have a single way to do these adjustments. “It's fairly complicated, because what the proper adjustments are is not so obvious,” he says. VaxGen originally claimed a P value of less than 0.02 for the black subgroup, meaning the company had a 98% or greater confidence that the result was not due to chance. Biologists typically use a P value of 0.05, or 95% confidence, as the dividing line between statistical significance and insignificance. Korber and Self argue that a widely used adjustment known as the Bonferroni correction should be applied to these results. The correction simply multiplies the P value by the number of subgroup analyses conducted. Gurwith says VaxGen did nine substudies based on race. A Bonferroni correction would change the P value for the black subgroup to between 0.09 and 0.18. “So it wouldn't be significant,” acknowledges Gurwith. He says the finding of significance in the group that combined blacks, Asians, and people of mixed race would remain, however. (Adjusted, P would be less than 0.04.) “This looks like a real result, and it makes some biological sense,” he says, noting that preliminary analyses show a correlation between anti-HIV antibody levels in vaccinated people and protection from HIV infection. Cornell University's John Moore, a longtime critic of the vaccine and an expert on HIV antibodies, finds this reasoning absurd. “Blacks and Asians lumped together is biological rubbish,” says Moore. “They might as well do a subgroup analysis based on signs of the zodiac.” And Self questions whether statistical significance holds up even in the combined group. “It's all murky because it's all post hoc analysis,” says Self. “There's some marginal effect [of the vaccine], and it's worth going after, but it's not worth overblowing. It's a hypothesis-generating result.” The confusion over VaxGen's results took another odd turn on 27 February. That afternoon, VaxGen CEO Lance Gordon told an investor conference in New York City that not only had the company done a Bonferroni analysis, but “a conservative version was applied, and it had no impact on statistical significance. … These are the accurate estimates, and the P values stand, even in view of subgroup and multiple subgroups.” He asserted that the company had “proven” that its vaccine “does prevent HIV at least in those populations most responsive.” Later that day, VaxGen issued a terse press release, noting that the company had followed the analysis plan they had agreed on with the U.S. Food and Drug Administration. The number of necessary adjustments remains “subject to interpretation,” the release says. “The company cannot predict the impact these adjustments may have on the findings, since that determination will ultimately rest with regulatory authorities.” Anthony Fauci, head of the National Institute of Allergy and Infectious Diseases, says he hopes the confusion will be cleared up at an AIDS meeting in Banff, Canada, that starts on 29 March: “Why don't we do what scientists always do, and settle this once and for all in an open forum?” 3. PLANETARY SCIENCE # Researchers Watch Solar Eruption Sideswipe Bright Comet 1. Robert Irion A storm from the sun has struck a glancing blow on the tail of a passing comet, thrilling astronomers and solar physicists alike. The encounter with comet NEAT, captured by a solar satellite, should yield new insights about the twisted magnetic churnings of the sun's outbursts. “This is a fantastic laboratory for studying the interaction between the sun and a violently active comet,” says planetary scientist Carey Lisse of the University of Maryland, College Park. Formally known as C/2002 V1, comet NEAT was spotted in November by NASA's Near Earth Asteroid Tracking (NEAT) system in Hawaii. It grew bright enough in early February for viewers in the Northern Hemisphere to see it after sunset with binoculars. On 18 February, the comet dove within 14.8 million kilometers of the sun, about 1/4 the distance of Mercury's orbit. That put NEAT within viewing range of a camera aboard the Solar and Heliospheric Observatory (SOHO), a joint NASA-European Space Agency satellite launched in 1995. As its icy surface roasted at about 1000 kelvin, the comet disgorged a surprisingly broad dust tail and a fainter tail of ionized gas. NEAT became the brightest of more than 600 comets seen by SOHO's Large Angle and Spectrometric Coronagraph (LASCO) instrument, which blocks the sun's light to expose its blazing outer atmosphere and its steady solar wind of charged particles. When the comet's value became clear, LASCO operator Kevin Schenk fought through knee-deep snow to reach the closed operations center at NASA's Goddard Space Flight Center in Greenbelt, Maryland, to send special instructions to the satellite. Notably, shorter exposure times revealed features otherwise lost in the comet's glare. SOHO's photos show the solar wind bombarding NEAT in unprecedented detail, says LASCO operations scientist Gareth Lawrence. In particular, the sun unleashed two giant eruptions of gas and tangled magnetic fields, called coronal mass ejections, near the comet as it whipped past at 100 kilometers per second. Researchers studying the images concluded last week that the edges of one ejection scrambled the ionized plasma tail, an event never before seen. “The tail seemed to split and form a wishbone,” says solar physicist Jeff Morrill of the Naval Research Laboratory in Washington, D.C. One part of the ion tail stayed with the moving head of the comet while the rest rippled along the margins of the dust tail, as if a magnetic structure within the solar ejection held it in place, Morrill says. The same eruption may have swept part of the dust tail into space, something else astronomers had not previously imaged. The comet's vigorous display also promises to expose much about its own history, says Lisse. The wide dust tail reveals that flecks of many different sizes boiled off NEAT's nucleus, because larger particles are less affected by the solar wind's drag. “It's a fantastically structured and highly spread out tail, the best I've seen since comet West in 1975,” Lisse says. As the comet recedes on its 37,000-year orbit, Lisse and his colleagues hope those grain sizes will help them reconstruct the conditions of NEAT's icy birth. 4. FORMER SOVIET UNION # Fears Grow of Nuclear Brain Drain to Iran 1. Richard Stone CAMBRIDGE, U.K.—It has long been a nightmare scenario for Western observers: so-called rogue nations luring impoverished nuclear experts from ex-Soviet republics to help them develop weapons. Now Eduard Shevardnadze, president of the Republic of Georgia, is putting some flesh on the fears. At a press conference last month, Shevardnadze claimed that several nuclear physicists from a breakaway region in the Caucasus are working in Iran. Western experts are also worried about the security of nuclear materials in the region. A nonproliferation official at the U.S. Department of Energy calls Shevardnadze's remarks “credible,” adding “it is one of our biggest fears.” Other analysts want to see more evidence but acknowledge that the lab and the region the physicists hail from—the Sukhumi Institute of Physics and Technology (SIPT) in Abkhazia, an isolated republic on the Black Sea—are a long-standing proliferation headache. In its Soviet heyday, SIPT had a 1500-strong staff and did everything from basic nuclear physics to applied research on radiothermal generators, devices that can power satellites by harnessing the heat from radioactive isotopes. All that work ground to a halt during a civil war in the early 1990s as Abkhaz separatists fought for independence from Georgia. About 200 ethnic Georgians were evacuated from the institute in September 1993. A Georgian official says his government has given U.S. officials the names of seven SIPT physicists who it claims are now working in Iran. The official declined to comment on whether the scientists have weapons knowledge or what precisely they are doing, apart from “teaching and research.” Some analysts, however, are withholding judgment. In light of the ongoing Georgian-Abkhaz conflict, says Scott Parrish, an expert on nuclear proliferation at the Monterey Institute of International Studies in California, “Shevardnadze has every motivation to accuse Abkhazia of various nefarious activities. … I'd certainly like to see some evidence from a less biased source.” Loose nuclear cannons are not the only fear. During the chaotic evacuation of Georgian scientists, or shortly thereafter, SIPT's entire stash of weapons-grade highly enriched uranium (HEU) apparently disappeared. Last year, Parrish met with Walter Kashiya, SIPT's “director-in-exile” in Georgia's capital, Tbilisi. Kashiya told him that SIPT's HEU—as much as 2 kilograms—had been used in experiments connected with the radiothermal generators. “By all accounts, there is now no HEU left in Sukhumi, and no good information on what happened to the material,” Parrish says. Some experts believe that 2 kilograms of HEU is enough to fashion a crude nuclear device. Other radioactive materials are also causing concern. Abkhazia, only slightly bigger than the U.S. state of Delaware, has a variety of materials that could, in principle, be used to make radiological dispersal devices, or “dirty bombs.” Most of Abkhazia's sources are “minor and appear to be under control,” says Abel Julio González of the International Atomic Energy Agency in Vienna, who was part of a delegation that visited the republic last year to discuss the problem. However, Soso Kakushadze, the head of Georgia's Nuclear and Radiation Safety Service, says one source is worrisome: The Institute of Experimental Pathology and Therapy in Sukhumi has radioactive cesium chloride in an easily dispersible powder form. The cesium, encapsulated in lead, was used to irradiate monkeys and other animals to study the health effects of radiation exposure in space. One capsule was stolen last spring and reportedly recovered in October. The apparent vulnerability of the pathology center and SIPT represents “a very dangerous situation,” asserts Zurab Tavartkiladze, first deputy at Georgia's environment ministry. “We don't know what will happen in Abkhazia tomorrow.” 5. CANCER RISK # Review Rules Out Abortion-Cancer Link 1. Jennifer Couzin National Cancer Institute (NCI) Director Andrew von Eschenbach doesn't have much room to maneuver. Last June, 28 abortion opponents in Congress complained about a fact sheet on NCI's Web site that reported no association between abortion and an increased risk of breast cancer (Science, 12 July 2002, p. 171). Shortly thereafter, NCI yanked the fact sheet and summoned 100 cancer experts, ranging from epidemiologists to mouse modelers, to a workshop* in Bethesda, Maryland, last week to review the evidence. Their conclusion: There is no association between abortion and breast cancer. This week, NCI's board of scientific advisers unanimously accepted the report from the workshop's participants and passed it on to von Eschenbach. He must now decide how to handle this politically volatile topic. An institute spokesperson says that NCI will consider putting an updated information sheet back up on its Web site. The meeting participants had an animated discussion of miscarriage, lactation, and other elements of pregnancy and sought to explore how they affect breast tissue and a woman's likelihood of developing breast cancer. But they turned up no significant new insights on the topic that had brought them there in the first place. Although one researcher, endocrinologist Joel Brind of Baruch College in New York City, argued in favor of an abortion-cancer link, the other attendees agreed that scientific research does not indicate excess risk. Several epidemiologists presented unpublished work, which built on existing studies, that found no abortion-breast cancer connection. One, led by Leslie Bernstein of the University of Southern California in Los Angeles, is following more than 9000 women. It has found “no evidence of an increase in risk.” Bernstein reported that miscarriage, although not abortion, appears to confer some protection against breast cancer, which full-term births are already known to do. She hypothesizes that human chorionic gonadotropin, a hormone that peaks 8 weeks into pregnancy before tapering off, may be the protective agent. If that's correct, she asks, “how can induced abortion be a problem” for cancer risk? • *”Early Reproductive Events and Breast Cancer,” 24–26 February 2003. 6. INFECTIOUS DISEASE # Shortage of Meningitis Vaccine Forces Triage in Burkina Faso 1. Gretchen Vogel Despite pleas last fall to prepare for a serious outbreak of bacterial meningitis this winter, public health officials are scrambling to halt a mounting epidemic in Burkina Faso, caused in part by an uncommon and hard-to-fight strain of the Neisseria meningitidis bacterium known as W135. Although some supplies of a new vaccine effective against W135 reached the West African country last week, they are not nearly enough, says William Perea of the World Health Organization (WHO). Public health officials are now forced to ration the limited supply in a way that saves as many lives as possible. Less well known than AIDS, tuberculosis, and malaria, bacterial meningitis kills and disables thousands of victims in the so-called meningitis belt of Africa each year. Transmitted through airborne droplets, it grabs hold in the winter as dry winds blow south of the Sahara and leave irritated airways vulnerable to infection. In last year's outbreak, 1500 people died and hundreds more were left with brain damage and other permanent disabilities. With 3000 cases already, and more than 400 deaths, this year promises to be even worse, says Perea. Preventive vaccination is key, as the disease is not easy to treat. Even under ideal conditions in the United States or Europe, with access to advanced antibiotics, one in 10 patients die and up to 15% suffer permanent complications such as deafness, brain damage, or limb amputation. A relatively cheap vaccine can prevent the most common strains of the bacterium, the A and C strains, but last year's outbreak in Burkina Faso was caused by the W135 strain. At the time, the only vaccine effective against the W135 strain costs between$4 and $40 per dose—far out of reach of people in Burkina Faso and most other African countries. Fearing a repeat, WHO and other public experts and nongovernmental organizations urged drug companies to produce a less expensive vaccine in preparation for this year's meningitis season, which typically runs from December to June (Science, 11 October 2002, p. 339). They also asked wealthy countries to contribute$10 million to help buy enough doses to prevent an epidemic. The response from donors was tepid. Norway and Monaco offered small donations, but the Bill and Melinda Gates Foundation made the only significant contribution, says Graciela Diap of Doctors Without Borders, which is helping respond to the crisis. “We were trying to pre-position some vaccine stocks in the field to be prepared for the epidemic,” says Diap, but most donors said to come back once there was a crisis. The limited funding slowed the negotiations, but early this year, drug giant GlaxoSmithKline (GSK) agreed to produce a cheaper variant of its existing vaccine, effective against three strains of the bacterium.

By that time, the first cases of meningitis—caused by both the W135 and the A strains—had been reported in Burkina Faso. After a fast-track approval in Belgium, the first 500,000 doses of GSK's trivalent vaccine were shipped to Burkina Faso last week. The country has requested 1.9 million doses.

WHO experts are now trying to figure out how best to distribute the available doses of vaccine. With limited data, they are trying to assess quickly which strains are prevalent where—the A strain can be treated with a cheaper vaccine—and deliver the precious trivalent vaccine to areas where the W135 strain is most threatening.

The crisis highlights an old debate among meningitis experts. John Robbins of the National Institute of Child Health and Human Development in Bethesda, Maryland, argues that routine vaccination, rather than emergency vaccination during a crisis, could prevent epidemics and could also reduce the cost of the vaccines by providing a guaranteed market. Several countries, including China, have implemented routine strategies and no longer suffer epidemics as they did through the 1970s, he argues.

But WHO officials and other experts assert that the benefits of routine immunization with existing vaccines are not worth the financial and logistical costs. In fact, one study published in January 2000 in The Lancet calculated that an organized response to an outbreak saved approximately the same number of lives as did a preventive vaccination campaign.

As the latest outbreak in Burkina Faso illustrates, however, launching an organized response is easier said than done. Both routine vaccination and the outbreak-response strategy face significant hurdles, notes Mark Miller of the National Institutes of Health Fogarty International Center, but he believes that routine vaccination is worth exploring. “I would think it would be very useful to do a demonstration project” in a region at high risk for outbreaks, he says. “The current strategy deserves a reassessment.”

7. ENVIRONMENTAL PROTECTION AGENCY

1. Rebecca Renner*
1. Rebecca Renner is a writer in Williamsport, Pennsylvania.

Geochemist James Kubicki didn't like it when the U.S. Environmental Protection Agency (EPA) rejected his proposal last fall to develop a novel way to detect the presence of any harmful fluoride compounds in drinking water. But his unhappiness turned to rage when EPA informed him a few months later that it had approved a request for a copy of his proposal. “I had no idea that they could do this,” says an indignant Kubicki, a professor at Pennsylvania State University, University Park. “Why should someone have the advantage of something that took my colleague and me weeks of work to prepare?”

The same thing happened to Michael Morris, an analytical chemist at the University of Michigan, Ann Arbor, although the blow was softened by the \$100,000 grant Morris received to tackle the same problem. “I usually get grants from the National Institutes of Health [NIH], where the proposals are confidential,” he says. “I was surprised to find that EPA's policy differs from [those of] other agencies.”

Morris's concern is well founded. But it's not the whole story. EPA's decision to grant such requests not only puts the agency at odds with other federal research agencies, it has also sparked internecine warfare. The requests for Kubicki's and Morris's submissions were approved by officials at the National Risk Management Research Laboratory in Cincinnati, Ohio, one of five divisions within EPA's Office of Research and Development, which had asked scientists for their ideas on monitoring drinking water. The lab acted after receiving guidance from EPA's Office of General Counsel in Washington, D.C.

But the director of another EPA division, one that handles the bulk of the agency's competitive grants, says that he would never accede to such a request. “We do not under any circumstances release unfunded proposals or grant applications, because these proposals represent private intellectual property,” says Peter Preuss, director of EPA's National Center for Environmental Research.

No one at EPA seems to have been aware of the discrepancy until last week, when Science began to make inquiries. Despite a flurry of internal communications, the agency's policy remained unclear at press time. And that leaves Myron Coplan in limbo.

Coplan, a chemical engineer and consultant at Intellequity in Natick, Massachusetts, believes that fluoridation is detrimental to public health. Last November, he asked the Ohio lab for all the proposals under its initiative, citing the Freedom of Information Act (FOIA), a 35-year-old law that requires the federal government to disclose certain records. The general counsel's office advised the lab that unfunded proposals must be released unless the information in them is covered by exemptions in the law. One commonly cited exemption covers personal privacy, such as an applicant's Social Security number or home address, and another covers business information. But the general counsel's office, citing a 1974 decision by the U.S. Court of Appeals for the D.C. Circuit (Washington Research Project v. HEW), advised the Ohio lab that this exemption does not cover novel ideas, research techniques, or methods in a research proposal unless they possess obvious commercial value.

Ironically, some lawyers who follow privacy issues say that the lab's decision to release the proposals is the right one. FOIA's exemptions are narrowly drawn to prevent an agency from making a blanket exception for all unfunded proposals, says Daniel Stotter, a FOIA attorney in Eugene, Oregon. David C. Vladeck, associate professor of law at Georgetown University Law Center in Washington, D.C., agrees: “It would be very difficult to challenge [the general counsel's position]” that exemption 4, which covers trade secrets and financial or commercial information, applies to grant applications.

Other agencies see it differently. The National Science Foundation can keep grant proposals confidential, says its FOIA officer, Leslie Jensen, because it stores the information in a records system that imposes many more restrictions on access under the federal Privacy Act. NIH's policies also prohibit the release of unfunded proposals, says FOIA officer Susan Cornell, on the grounds that there is enough private information and commercial value in every proposal. Both officials say that funded proposals are another matter, however, and are likely to be released under FOIA.

Kubicki had never heard about FOIA when he submitted his proposal, and he doesn't understand why some government officials think his ideas should be in the public domain. “I proposed something fairly novel in that grant. I would hate to think that this [law] is a way for people to get ideas from other researchers.”

8. INFECTIOUS DISEASE

# Chasing the Fickle Swine Flu

1. Bernice Wuethrich*
1. Bernice Wuethrich is a science writer in the Washington, D.C., area.

Changes on the farm may be fostering the evolution of the swine flu virus—and if a dangerous new strain crosses back into humans, it could have deadly consequences

One of the first signs of trouble was a barking cough that resounded through a North Carolina farm in August 1998. Every pig in an operation of 2400 animals sickened, with symptoms similar to those caused by the human flu: high fever, poor appetite, and lethargy. Pregnant sows were hit hardest, and almost 10% aborted their litters, says veterinary virologist Gene Erickson of the Rollins Animal Disease Diagnostic Laboratory in Raleigh. Many piglets that survived in utero were later born small and weak, and some 50 sows died.

The culprit, a new strain of swine influenza to which the animals had little immunity, left veterinarians and virologists alike puzzled. Although related flu strains in birds, humans, and pigs outside North America constantly evolve, only one influenza subtype had sickened North American pigs since 1930. That spell was suddenly broken about 4 years ago, and a quick succession of new flu viruses has been sweeping through North America's 100 million pigs ever since. This winter, for example, up to 15% of the 4- to 7-week-old piglets on a large Minnesota farm died, even though their mothers had been vaccinated against swine flu, says veterinary pathologist Kurt Rossow of the University of Minnesota, Twin Cities.

It seems that after years of stability, the North American swine flu virus has jumped onto an evolutionary fast track, churning out variants every year. Changes in animal husbandry, including increased vaccination, may be spurring this evolutionary surge. And researchers say that the resulting slew of dramatically different swine flu viruses could spell danger for humans, too. The evolving swine flu “increases the likelihood that a novel virus will arise that is transmissible among humans,” says Richard Webby, a molecular virologist at St. Jude Children's Research Hospital in Memphis, Tennessee.

Because people have no immunity to many viruses from other species, strains that on rare occasions leap the species barrier can have deadly consequences (see sidebar, p. 1504). And pigs are considered “mixing vessels” in which swine, avian, and human influenza viruses mix and match. Scientists believe, for example, that the last two flu pandemics, or worldwide epidemics, in 1957 and 1968, occurred when avian flu and human flu viruses swapped genes in pigs, creating a new, hybrid virus that then spread to humans. In each case, the new virus appeared first in Southeast Asia, then around the globe. The 1918 “Spanish flu,” which claimed upward of 40 million lives, may also have arisen first in pigs. “We used to think that the only important source of genetic change in swine influenza was in Southeast Asia,” says Christopher Olsen, a molecular virologist at the University of Wisconsin (UW), Madison. Now “we need to look in our own backyard for where the next pandemic may appear.”

Fortunately, the new pig strains that have appeared in North America so far do not appear to readily infect humans. But researchers are sufficiently concerned that they are calling for increased surveillance of both humans and pigs. “Within the swine population, we now have a mammalian-adapted virus that is extremely promiscuous,” says Webby, referring to the virus's proclivity to swap genes with avian and human flu influenza viruses. “We could end up with a dangerous virus.”

## When pigs fly

Most genetic changes in the flu viruses—human, pig, and bird—are small and subtle point mutations in the virus's RNA. Less common but more alarming are sudden, wholesale changes that replace entire genes and are more likely to circumvent the immune system. This process, called genetic shift, is exactly what is now occurring in North American pigs. Thus, the latest swine influenza virus is a curious hybrid: The genes that code for its coat proteins derive from classical swine influenza, but half of its internal genes have been snatched whole from avian and human viruses.

The structure of the influenza virus lends itself to such radical changes. The virus is made of eight single-stranded segments of RNA that together code for 10 proteins (see illustration). If two or more different viruses infect the same host cell, they can swap segments, creating new viral types.

Most commonly, the virus swaps genes that code for its two surface proteins: hemagglutinin (HA) and neuraminidase (NA). Both proteins spike off the virus's outer coat, and HA initiates infection when it binds to receptors on host cells. The immune system of the infected animal targets sites on these molecules. Therefore, a virus with a novel HA can escape preexisting immune defenses—hence the pig deaths.

Influenza viruses are named after their HA and NA components, as in “H1N1” or “H3N2.” Both human and swine influenzas have been historically limited to only a few of these varieties. Birds, on the other hand, can be infected by every combination of the virus's 15 HA genes and nine NA genes, forming a vast global reservoir of virus. And pigs have receptors for both human and bird flu virus, making them crock pots for new viral combinations.

The “classical” swine influenza virus discovered in 1931 is an H1N1 virus, related to the H1N1 that caused the 1918 pandemic. But in the past 5 years, a quick succession of progeny, which now include at least three additional virus subtypes and four genotypes, have all but supplanted that classical swine virus in North American pigs.

The first new virus, the one that struck the North Carolina hogs in 1998, was an H3N2; in this case, genes had crossed from human viruses to pig viruses. By late 1999, the novel viruses could be found wherever there were pigs in North America and so were presumably spread by cross-country transport. Webby and St. Jude colleague Robert Webster, together with Olsen and others at UW Madison, traced these viruses' evolutions. They found both “double reassortant” viruses, with human and swine flu genes, and “triple reassortants,” containing genes from human, swine, and avian influenza viruses.

The avian flu genes may hold clues to the viruses' evolution. They code for two of the virus's three polymerase proteins: PA and PB2. (A third polymerase, dubbed PB1, comes from either human or avian viruses, depending on the swine flu subtype.) All three polymerases are involved in viral replication, and they tend to do a sloppy job, allowing countless errors to slip by.

Scientists suspect that these three imported polymerase genes form a viral platform that Webby calls “unnervingly adept” at triggering change in the influenza genome. In fact, H3N2 has continued to change, acquiring a succession of HA genes derived from the human influenza viruses that circulated several years previously. By 2000, Olsen's lab had identified a new viral subtype, an H1N2 that is a combination of the classical swine virus and the H3N2.

This season's variant is an H1N1 with the internal genes of an H3N2. Its HA gene, derived from the classical swine influenza virus, appears to be rapidly mutating. The amount of sequence divergence among certain 2001 isolates “is as much as the difference between classical H1N1 viruses isolated in the 1960s and those isolated in the early 1990s,” Webby reports. If enough point mutations accumulate, that HA molecule could become unrecognizable to the immune systems of both pigs and humans.

## Pig pressures

As they seek to understand what upset the status quo in North American swine, researchers have turned to the environment. For North American pigs, the environment has recently changed dramatically in two ways: herd size and vaccination practices.

In the past decade, big swine producers have gotten bigger, and many small producers have gone out of business. The percentage of farms with 5000 or more animals surged from 18% in 1993 to 53% in 2002, according to Rodger Ott, an agricultural statistician at the National Agricultural Statistics Service in Washington, D.C. Having more pigs under one roof makes it more likely that a rogue virus can take hold. “With a group of 5000 animals, if a novel virus shows up, it will have more opportunity to replicate and potentially spread than in a group of 100 pigs on a small farm,” Rossow says. On the other hand, pigs in outside pens, as is common on small farms, can be exposed to the droppings of migratory waterfowl, which may contain infectious viruses; large-scale confinement agriculture may prevent such exposure, points out Liz Wagstrom, director of veterinary science at the National Pork Board in Clive, Iowa.

Another crucial change has been the recent wide-scale vaccination for swine influenza. In less than a decade, vaccination has become the norm for breeding sows, which in turn pass their maternal antibodies on to their progeny. In 1995, swine flu vaccination was so new that the National Swine Survey conducted by the United States Department of Agriculture didn't bother to assess its extent. In 2000, the same survey showed that 44.1% of breeding females received a vaccine. Today, more than half of all sows are vaccinated against both H1N1 and H3N2 viruses, says Robyn Fleck, a veterinarian at Schering-Plough, one of the nation's three producers of swine influenza vaccine. But the vaccine is not protecting against all new strains. “We're seeing clinical disease in vaccinated pigs,” says Rossow. Flu is also showing up in piglets thought to be protected by maternal antibodies passed on from vaccinated sows, such as those on the Minnesota farm.

Widespread vaccination may actually be selecting for new viral types. If vaccination develops populations with uniform immunity to certain virus genotypes, say H1N1 and H3N2, then other viral mutants would be favored. Webby suggests that the combination of avian polymerase genes generating errors in the genetic sequence and immunologic pressure from vaccination may be selecting for unique variants. However, he adds that “the benefits of vaccination far outweigh this side effect.” From a human perspective, reducing the overall viral burden in pigs through vaccination is a plus. “If you can decrease the amount of virus, you can reduce the chances of interspecies transmission,” he says.

Schering-Plough veterinarian Terri Wasmoen acknowledges that vaccines “may be pressuring change.” But she also notes that larger hog confinement operations and more shipping from state to state may play a role. “We need epidemiological work to understand these issues, and there is no funding now,” she says.

## Breaking the species barrier

Although scientists know that influenza viruses can jump the species barrier directly, such an event has been seen as a rarity. A review of the literature yields only 18 cases of pure swine influenza directly crossing into people, says Olsen. But his work suggests that many cases of transmission may occur, then fizzle out. Olsen tested 74 swine farm owners, employees, their family members, and veterinarians in rural Wisconsin for antibodies to swine influenza and compared the results to those for 114 city folk in a study published last summer in Emerging Infectious Disease. Seventeen of those routinely exposed to pigs tested positive for the antibodies, whereas only one urban dweller did so.

Olsen and his colleagues have also found evidence that a novel H4N6 swine virus isolated in pigs in Ontario—which probably came from local ducks—has already acquired genetic mutations that give it the potential to bind to human cell receptors. Such an event could be catastrophic, as humans have no immunity to H4 viruses. But getting into humans is just the first step. To have pandemic potential, a new influenza virus must also be able to move easily from one person to the next. No new virus from swine or birds, nor any hybrid created in pigs, has been able to accomplish this since the 1968 pandemic.

Even so, experts in both animal and human health are beginning to call for increased surveillance to stop a new pandemic before it starts. The World Health Organization (WHO) constantly scans the globe for new strains of human influenza, which are used to make annual recommendations for next year's vaccines. But “there is no systematic monitoring of [human] populations where there may be interspecies transmission between humans, birds, and pigs,” says Carolyn Buxton Bridges, an epidemiologist at the U.S. Centers for Disease Control and Prevention in Atlanta.

In addition, “we don't have any official surveillance system for swine influenza,” says Sabrina Swenson, head of bovine and porcine viruses at the National Veterinary Services Laboratories in Ames, Iowa. “We have to bring the human-health people together with the veterinary-health people because of concern that the viruses can move to people. It would be nice to have something better defined, but it's dependent on funding,” she says.

Webster, who heads the WHO collaborating laboratory on animal influenza in Memphis, Tennessee, calls for the development of reagents to recognize every possible influenza subtype and for a global network for monitoring of animal influenzas at the human-animal interface. “If something strange pops up in Georgia or Washington state, do we have the reagents in place to move quickly from identification to making a human vaccine?” he asks.

Last month, at a WHO meeting to determine next year's vaccine constituents, Webster presented his data on animal influenza and urged the development of vaccines for every HA subtype, to be used as a first defense should an unusual virus appear in the human population. WHO agreed and is developing a prioritized list of viral subtypes that it will send to all vaccine manufacturers, urging preparedness for the next pandemic. “It wouldn't provide total protection, but it could keep the virus from killing you,” Webster says.

9. INFECTIOUS DISEASE

# An Avian Flu Jumps to People

1. Bernice Wuethrich*
1. Bernice Wuethrich is a science writer in the Washington, D.C., area.

While researchers monitor U.S. pigs for potentially dangerous changes in swine influenza virus (see main text), recent events on the other side of the world have sounded an even more urgent alarm. Last month in Hong Kong, a 33-year-old man died and his 9-year-old son fell seriously ill after contracting an avian influenza virus from a source that remains mysterious.

Initial genetic sequencing suggests that the virus may be descended from one found in wild birds. If so, it could be difficult to contain. In all previously known cases of an avian flu jumping to humans, the source is believed to have been poultry. But “this virus hasn't been seen in domestic poultry,” says Robert Webster, director of the World Health Organization's (WHO's) collaborating laboratory on animal influenza at St. Jude Children's Research Hospital in Memphis, Tennessee. However, authorities have not ruled out the possibility that the virus came from chickens on a relative's farm in mainland China.

Most flu viruses are adapted for a particular group of animals, although pigs can mix and match viruses from birds and humans. And on seemingly rare occasions, flu viruses have jumped the species barrier from other animals to humans. The last two human influenza pandemics, or worldwide flu epidemics, were caused by viruses that incorporated both human and avian flu genes. Because humans have no immunity to many strains of avian influenza, such viruses have deadly potential if they acquire the ability to infect human cells and move easily from one human host to another. However, Hong Kong health officials say that thus far there is no indication that the latest avian virus can spread from person to person.

All the same, WHO has declared an influenza alert, and its collaborating laboratories have moved into high gear. “We don't know what the consequence of this virus will be. I wouldn't trust it,” says Webster. He points out that in the most similar case known—an avian influenza virus that jumped from chickens to humans in 1997, infecting 18 people in Hong Kong and killing six—there was a 6-month lag between the first death and those that followed.

Further complicating the current case is the fact that the Hong Kong family had recently visited relatives who keep chickens in Fujian Province in mainland China. The man's 8-year-old daughter died from an undiagnosed respiratory infection while there. And the deaths in this family coincide with an unidentified virus sweeping nearby Guangdong Province, infecting more than 500 people and killing at least seven. Researchers aren't sure if the family has the same illness as those in Guangdong. Webster says the nature of the Guangdong outbreak is still unknown, but “WHO has a team investigating there now.”

The virus in the 1997 avian flu outbreak, labeled H5N1 for the particular forms of its surface molecules (“H” stands for hemagglutinin and “N” for neuraminidase), didn't spread beyond those who'd had direct contact with infected birds. All 1.4 million chickens in Hong Kong were slaughtered with the hope of stamping out the virus (Science, 16 January 1998, pp. 324 and 393). But despite the massive slaughter, the virus survived—perhaps in backyard poultry or in other domestic or wild bird hosts—and has continued to evolve. The latest virus, also an H5N1 subtype, is believed to be its most recent descendant.

The HA molecule of the new virus is most similar to that of the 1997 H5N1, says Klaus Stöhr, director of the WHO influenza program in Geneva, but its overall genetic makeup most resembles that of a recently discovered relative that infected migratory wildfowl. That H5N1 virus felled ducks, geese, swans, greater flamingos, and wild little egrets in a Hong Kong park beginning last November. This means that the deadly H5N1 may already have spread. “It suggests [that] the virus is not necessarily confined to that park in Hong Kong, because it was found in free-roaming waterfowl,” Stöhr says.

It is unusual to see waterfowl sick with the flu, Stöhr says. Avian influenza viruses usually bind to receptors in the intestinal tract of their wild bird hosts and cause no symptoms. “The surprise was to find a highly pathogenic virus in waterfowl—the natural host for influenza,” says Stöhr. Domestic poultry do harbor low-pathogenic influenza strains. A highly pathogenic H5N1 virus, such as the 1997 Hong Kong virus, typically becomes deadly by mutating so that it can bind to cells throughout the bird's body, rather than just in the intestinal tract. Then it can cause systemic and fatal disease in domestic poultry such as chickens and turkeys, explains Terrence Tumpey, an avian influenza expert at the U.S. Department of Agriculture's Southeast Poultry Research Laboratory in Athens, Georgia. No one knows how such a virus might acquire the ability to infect humans, however, as avian influenza viruses typically bind to cell receptors not known to exist in people.

Now that another H5N1 influenza virus has crossed into the human population, public health experts are taking no chances. All the chickens in Hong Kong are undergoing vaccination for the virus, and a large number of already infected or potentially infected birds are being slaughtered. The park where the infected wild and exotic birds died was closed and disinfected, and the wild birds were killed.

WHO is helping coordinate preliminary work on a vaccine, e-mailing the viral sequence to its collaborating labs around the world. However, the virus is so virulent that it kills chick embryos within 2 to 3 days, making it impossible to use eggs to produce the vaccine, as is usually done, Stöhr says. Rather, labs will have to use the viral sequence to create a less virulent form of the virus in tissue culture, and then put this virus into eggs for vaccine production. “We are desperately working on a vaccine through reverse genetics,” says Webster.

“The goal is to make available a vaccine strain to all vaccine companies,” Stöhr adds—and so to have a vaccine available should this species-hopping virus manage an even more alarming leap: from one human host to the next.

10. HIV

# Escape Artist Par Excellence

1. Jon Cohen

Recent findings are giving researchers new respect for HIV's Darwinian prowess in evading the body's immune defenses

BOSTON, MASSACHUSETTS—Move over, Darwin's finches. As examples of evolution in action, few organisms are better than the AIDS virus. Researchers have known for nearly 20 years that the virus gains a huge evolutionary advantage by targeting immune system cells that the body sends out to defeat it. But that's just its most obvious trait. New studies presented here last month during the largest annual U.S. AIDS conference* provide fresh insights into how the virus has deftly adapted to humans, hijacking, mimicking, and dodging defenses that otherwise might keep it at bay.

Few of these new findings captured headlines from the meeting, and some immediately kicked up controversies. But they prompted a buzz among AIDS researchers here, many of whom have been trying to take the measure of this virus for 2 decades. Every new piece of evidence seems to produce new respect for HIV's Darwinian prowess. And some of these new insights may also uncover weaknesses in the virus's highly evolved modus operandi. “Some years from now, people will look back and say ‘These are the breakthroughs,’” predicts David Ho, the chair of the meeting, who heads the Aaron Diamond AIDS Research Center in New York City.

HIV apparently has picked up clever tricks to get around an arm of the immune system known as innate immunity. Unlike the more refined adaptive immunity, which relies on white blood cells that learn to recognize foreigners and then attack them with exquisite precision, innate immunity uses “natural killer cells,” biochemicals, and cellular proteins that behave with more brawn than brain. They take out whole classes of invaders with little discrimination and no memory of what they've done.

At the meeting, Paul Bieniasz, a virologist who works at Aaron Diamond, reported that in human cells, the virus evades a key mechanism of innate immunity that targets retroviruses, the family that includes HIV, by hijacking a critical protein. “It's a very interesting story,” says Joseph Sodroski, a virologist at Boston's Dana-Farber Cancer Institute whose lab laid the foundation for Bieniasz's study.

Bieniasz became intrigued by the Sodroski group's discovery that HIV could not copy itself in most monkey cells, even when the researchers engineered the virus so that it could enter them. Bieniasz, Greg Towers of University College London, and their co-workers wondered whether the monkey cells contained something that inhibited the virus or, alternatively, whether human cells had something that aided it. They tried fusing the monkey cells to human cells, but HIV still could not replicate in the hybrid cell, strongly suggesting that the monkey cells harbored an effective inhibitor.

Bieniasz and his colleagues discovered that this inhibitor cripples HIV's ability to copy itself by binding to a protein, known as capsid, that forms a wall around the virus's genetic material. In human cells, they found, HIV likely evades this inhibitor, dubbed Restriction factor-1, by binding a human cellular protein called cyclophilin to its capsid, blocking Restriction factor-1.

A study reported in Nature last summer (Science, 19 July 2002, p. 312) by Michael Malim's group at King's College London indicated that HIV had found a way to defeat another barrier in human cells. Malim's group found that HIV's Vif protein blocked an innate viral inhibitor called CEM-15. “When you view it in light of what Malim found, these naturally occurring resistances to retroviruses are probably a lot more widespread and important than we thought,” says Bieniasz. But HIV has evolved its way around them.

## Avoiding antibodies

When it comes to adaptive immune responses to HIV, antibodies have received much of the attention—and caused unparalleled frustration and confusion. Antibodies typically derail viruses by binding to specific parts of their surface proteins, preventing the invaders from locking onto receptors on cell surfaces, the first step in establishing an infection. HIV long has befuddled AIDS vaccine developers because it overcomes almost every antibody that attaches to its surface protein, gp120. The central genius of the virus is that it copies rapidly and mutates frequently, creating a staggering number of viral strains. But over the years, researchers have fished a half-dozen antibodies from infected people that are remarkably potent—and work against many strains of the virus. Although no research group has yet found a protein that can stimulate the body to produce such “broadly neutralizing” antibodies, investigators hope that a finer understanding of them will lead to the Holy Grail.

At the meeting, both Sodroski and Dennis Burton of the Scripps Research Institute in La Jolla, California, described some rare and potent antibodies that they have intensely studied and provided detailed models of how they do their magic. Each of these unusual antibodies has unusual structural features. “They all have something funny about them,” says Sodroski. One champion called 2G12, for example, looks more like an “I” than the standard “Y” shape of antibodies, whereas much-celebrated antibodies known as b12 and 17b both have long fingers that allow them to reach otherwise protected parts of gp120.

To Sodroski, this helps explain why AIDS vaccine researchers have had such disappointing results with vaccines that feature gp120, such as the one made by VaxGen of Brisbane, California, that last week received wide attention for failing in a full-scale efficacy trial (see p. 1495). Most experimental AIDS vaccines tested in humans to date, VaxGen's included, have relied on what amounts to garden-variety gp120 to stimulate antibodies. But in nature, gp120s are engaged in a kind of arms race with antibodies, always making new forms. The gp120s that trigger the production of exotic antibodies such as 2G12, 17b, and b12 therefore presumably bear little resemblance to the garden-variety versions.

Burton, Sodroski, and other researchers remain convinced that if they can find or engineer these unusual gp120s, they can stop HIV before it can get a foothold in the body. Their confidence comes in part from several “passive” immunization studies—including one with b12 by Burton and Cornell University's John Moore, published in the 10 February online edition of Nature Medicine—which demonstrate that directly administering antibodies to animals can, at least for a short time, thwart infection with the AIDS virus.

But on a more sobering note, recent evidence also makes it abundantly clear that once an infection is established, antibodies do little more than spray a garden hose on a forest fire. Douglas Richman, a virologist at the University of California, San Diego, drives this point home in a recent study done by his lab that analyzed the evolution of antibodies in 14 people shortly after they became infected with HIV, following them for up to 39 months.

In a paper in press at Proceedings of the National Academy of Sciences, Richman and co-workers report that HIV mutates to escape antibodies at a furious rate—much more quickly than it does to escape the pressure of anti-HIV drugs. “The rate of viral evolution because of antibody is at least an order of magnitude faster than drug-resistance develops,” Richman says. And although drug-resistant HIV mutants pay an evolutionary “penalty”—they cannot copy themselves as easily as “wild-type” viruses do and thus are not as fit—Richman's group found no such effect when antibody drives the selection process. “This virus is so pernicious,” says Richman.

## First principles

To better understand how HIV has evolved its armor against antibodies—and where the chinks might lie—Sodroski and, separately, James Hoxie of the University of Pennsylvania in Philadelphia, who also presented a paper at the meeting, have arrived at a revisionist theory about the virus and its relation to cellular receptors. Within a few years of discovering HIV, AIDS researchers learned that the virus initiates an infection of a white blood cell by docking gp120 onto a cellular receptor called CD4. The act of binding to CD4 causes the loop-shaped portions of gp120 to reorient themselves, as well as the sugars that form clouds around them. This “conformational change” exposes parts of the viral protein that until then had remained hidden. As researchers realized in 1996, these newly exposed portions of gp120, in turn, attach to a second “coreceptor” that normally binds immune system messengers called chemokines. This clicks open the final lock on the cellular door, and infection occurs.

After the discovery of HIV's relation to chemokine receptors, several investigators became excited by the idea that they had found a viral Achilles' heel: What if antibodies could attack gp120 after it had bound to CD4 and gone through its conformational change, but before it had attached to the chemokine receptor? A collaboration between the Burton and Sodroski labs shows, however, that CD4 yanks the virus close to the cell, physically preventing antibodies from reaching the newly exposed parts of gp120.

Indeed, Sodroski and Hoxie contend that the chemokine receptor, not CD4, was HIV's original point of entry into the cell; when the virus evolved to use the CD4 receptor, they argue, that resulted in added protection for the chemokine receptor binding site.

## Chemokine mimicry

Whereas most researchers have focused on how HIV's coat changes to disguise the virus from antibodies, immunologist Susan Zolla-Pazner of New York University has reported a remarkable evolutionary twist: A highly variable portion of the virus may be a surprisingly steadfast mimic of important immune system proteins.

Zolla-Pazner focused on the V3 loop, a hairpin-shaped projection from gp120 that researchers knew long ago triggers a particularly strong antibody response. Indeed, in the 1980s many leading vaccine designers believed that antibodies that were directed against this region of gp120 had unique power to stop HIV. But as researchers began to realize how much the loop varied—and that antibodies against it primarily worked on laboratory-grown, but not freshly isolated, strains of HIV—they concluded that a vaccine based on it likely would not have broadly neutralizing power. “Everyone dismissed the V3 idea as too strain specific,” recalls Richman.

Now Zolla-Pazner has identified a half-dozen new antibodies directed against the V3 loop that neutralize many strains of HIV. And she discovered a structural explanation that made her colleagues do double takes.

Working with Michal Sharon and colleagues at the Weizmann Institute of Science in Rehovot, Israel, Zolla-Pazner used nuclear magnetic resonance to determine the structure of the peptides in the V3 loop that bound to one of these broadly neutralizing antibodies. As the researchers detail in the February issue of Structure, when they compared the hairpin shape of their V3 loop peptides to structures in the Protein Data Bank, they found a striking homology to the very chemokines that bind to the receptors HIV uses to enter cells. “I was stunned,” says Zolla-Pazner. “It was one of the most amazing ‘aha’ experiences I've had in science. It's not something we were looking for.”

Although some researchers voiced reservations about the meaning of the structural similarities, Zolla-Pazner's finding made inherent sense to many: Because the V3 loop plays a critical role in helping HIV bind to chemokine receptors, it's logical that it would have the shape of the chemokines themselves. “There's a fundamental beauty in what she's talking about,” says Hoxie. “The themes of mimicry are very cool.”

Zolla-Pazner came up with another surprise: Although the amino acid composition of V3 peptides varies greatly, the loop retains its chemokine-like, hairpin shape. This may explain how antibodies that target a variable region of the virus could neutralize a broad range of viruses, until now a conundrum. “HIV uses many mechanisms that have evolved in the immune system against that very immune system,” concludes Zolla-Pazner. Talk about survival of the fittest.

• *10th Conference on Retroviruses and Opportunistic Infections, 10 to 14 February.

11. JAMES BATTEY PROFILE

# NIH's Man in the Middle of the Stem Cell Debate

1. Constance Holden

With researchers eager to push ahead and politicians worried about moving too fast, NIH's James Battey is trying to strike a balance on this controversial topic

James Battey knows that he works in a fishbowl. As head of a National Institutes of Health (NIH) panel managing stem cell research within the bounds established by the Bush Administration, Battey hopes to satisfy researchers who want to extend the scientific frontier without incurring the wrath of politicians who raise the specter of “human embryo farms.”

Last week, the House of Representatives reiterated its solid opposition to both reproductive and research cloning with human embryos. It voted for a total ban on nuclear transplants to create early human embryos, which the White House embraced immediately in hopes of persuading a divided Senate to follow suit. But the 241-155 tally in favor of a ban didn't rattle Battey. Six months into his stint as the government's point person on the issue, the 50-year-old cell biologist appears to be obeying his political masters without losing the respect of his peers. “He's the one who's trying to make the science work,” says stem cell researcher John Gearhart of Johns Hopkins University in Baltimore, Maryland. “I think he's in our corner, but it's a very narrow corner.”

Not many researchers are familiar with Battey, a 15-year NIH veteran whose full-time job is director of the National Institute on Deafness and Other Communication Disorders. Well regarded for his research on G protein-coupled receptors before moving into administration, Battey was tapped by NIH Director Elias Zerhouni to head up a task force that would implement the president's decision to limit federal support to stem cell lines derived before 9 August 2001 (Science, 17 August 2001, p. 1242). The 14-member internal panel (stemcelltaskforce.nih.gov) includes prominent stem cell researchers Ronald McKay and Mahendra Rao. It also receives advice from working groups of extramural grantees that include top researchers such as James Thomson of the University of Wisconsin, Madison, and Irving Weissman of Stanford University.

Science visited Battey at NIH last month during the area's seemingly endless winter to see how he's holding up. Blue-eyed and athletic, he looked as though he had just breezed into the office after a morning of snowshoeing. He was railing against the cumbersome and—he believes—ineffectual code orange security precautions at NIH in a tone that suggested a low tolerance for bureaucratic niceties. “Someone could drive a truck over a curb” or come in on the weekend without being hassled, he scoffed.

The first priority for the task force, Battey explains, is to beef up the number of researchers trained to work with human embryonic stem (ES) cells. “There are probably no more than 30 to 40 principal investigators,” he says. “I could see the field growing five- to 10-fold over the next 5 years.” To prime the pump, NIH is offering training funds to established investigators, short-term training courses for novices, and “administrative supplements” for researchers to add ES cells to their portfolio.

The next big step is to characterize the available cell lines. (Last November, Battey reduced the number of approved cell lines listed on the NIH registry from 74 to nine after weeding out those not available for proprietary reasons or because they hadn't yet been cultured.) Battey says he plans to advertise this spring for two researchers and two technicians to analyze the lines that are actually available to researchers. McKay, who will oversee the effort, says the aim is to provide “an information resource” for scientists seeking the best methods of cultivating particular cell lines. He says he has obtained six cell lines so far and expects to have 10 by the end of the year as more approved lines become available.

Another adviser to the task force, Leonard Zon of Children's Hospital in Boston, applauds this move, saying that “characterization of the lines as they currently stand is insufficient” for choosing which line to pick for a given experiment. “You don't even know which lines are genetically stable.” Adds Battey, “It's been assumed that cell populations grown from the same blastocyst are similar if not identical.” But detailed gene analysis might show them to be “very different.”

Characterization won't solve a deeper problem: There may be too few cell lines to satisfy NIH-funded researchers. Many would agree with Gordon Keller of Mount Sinai School of Medicine in New York City, another working group member, who thinks “additional lines will be necessary in the not-too-distant future” to provide greater genetic variation. Battey disagrees. “We need several years of research before we even need to think about more lines,” he says.

In the meantime, he sees NIH's “infrastructure” grants for ramping up stem cell production—eight groups in the United States and abroad have been funded to date—as a way to stimulate the field. He admits that some may take a while to bear fruit since using some lines that are currently frozen (including those at the Karolinska Institute in Sweden) may await new cultivation techniques.

Battey also takes a more optimistic stance than his nongovernmental colleagues on the potential effects of a ban on research cloning, should the Senate follow the House's lead. Research on diseases won't be severely hurt if the government blocks human cloning technology, he says: “I think we are many, many years away from having the knowledge we need to make such [alarmist] statements.” Battey argues that the most important good from human embryo cloning would be knowledge of how to reprogram a cell's nucleus—and “those kinds of lessons can most easily be learned in animal models.”

Despite their differing views, observers seem to feel that Battey is the man for the job. NIH and the task force are “doing as well as they can within the constraints,” says Keller. According to Zon, “What was needed was to have someone take charge and make sure stem cells were available to researchers.” By doing that, he says, Battey's task force is serving the community well.