# News this Week

Science  23 Jun 2000:
Vol. 288, Issue 5474, pp. 2106
1. SCIENTIFIC MISCONDUCT

# Panel Finds Scores of Suspect Papers in German Fraud Probe

1. Michael Hagmann

A new report paints a darker picture of what may be the highest profile case of scientific fraud in postwar Germany. After a 2-year investigation of all 347 scientific articles co- authored by former hematologist and cancer researcher Friedhelm Herrmann, a task force jointly sponsored by Germany's main granting agency, the DFG, and the country's largest cancer charity released a report on 19 June indicating that the scope of the fraud—first uncovered 3 years ago—is far more extensive than previously thought. Although Herrmann and a co-author have left their academic posts, the new revelations could place other careers in jeopardy.

A well-known and decorated member of the German research establishment, Herrmann studied various leukemias and how they respond to cytokines, signaling messengers of the immune system. The task force, headed by cell biologist Ulf Rapp of the University of Würzburg, concluded that 52 of Herrmann's papers “contain falsifications,” including data that were manipulated, while another 42 include data suspected of having been manipulated. The task force also found evidence of manipulation in three habilitations, the uniquely German post-Ph.D. qualification for aspiring professors, and “inconsistencies”—but no firm evidence of data falsification—in a randomly selected publication co-authored by clinician Roland Mertelsmann, chief of Herrmann's former hematology and oncology department at the University of Freiburg Medical Center. Any sanctions stemming from the report would come from the universities involved, the granting agencies that funded the fraudulent work, state and federal science ministries, and public prosecutors. “This is not the time yet to heave a sigh of relief and consider the matter closed,” says DFG Secretary-General Reinhard Grunwald.

View this table:

Initial allegations emerged in late January 1997, when Eberhardt Hildt, a postdoc in Herrmann's lab, claimed to his former Ph.D. supervisor that Herrmann, then at the University of Ulm, and Marion Brach, who had just left Ulm to start a lab at the University of Lübeck, had fabricated data in as many as four research papers appearing in 1994 and 1995. The two universities—as well as the University of Freiburg and the Max Delbrück Center for Molecular Medicine in Berlin, where Herrmann and Brach had worked previously—then set up panels to probe the allegations.

When questioned by the Lübeck panel, Brach said she was pressured by Herrmann to manipulate data on a few occasions in 1993 and 1994. Herrmann has steadfastly denied knowledge of or involvement in any manipulation. Meanwhile, the Berlin investigations zeroed in on autoradiograms—dark bands, resembling bar codes, whose patterns reflect protein and RNA production in cells under various experimental conditions. The panels found that individual bands and entire assemblies had been recycled to pose as results from other experiments in the same article or in later publications.

The local committees fed their findings to a national panel, chaired by Wolfgang Gerok, a retired professor of internal medicine. In a summary report issued in August 1997, Gerok's panel detailed alleged falsification and data manipulation in 37 publications. In the wake of the report, Herrmann and Brach resigned their positions.

The findings were “definitely a wake-up call for the German scientific community,” says Rapp. After much soul-searching, the DFG and the Max Planck Society, Germany's premier research institution, issued guidelines to guard against fraud that were subsequently adopted by most universities (Science, 19 December 1997, p. 2049). But the magnitude of the alleged misconduct prompted the panel to recommend, Gerok says, that all of Herrmann's publications “as well as his scientific environment be scrutinized.”

Beginning in April 1998, Rapp's four-member team located all of Herrmann's publications from the MEDLINE database. Because the three co-authors (besides Brach) with whom Herrmann published most frequently were all affiliated with Mertelsmann's hematology department, first at the University of Mainz and then at Freiburg, Rapp's team focused on work generated during Herrmann's years at both institutions, from 1988 to 1992.

The panel also sampled the work of other department researchers. “We wanted to determine … whether [fraudulent behavior] was restricted to Herrmann's group or typical for the entire department,” says Rapp. The task force examined more than 600 publications by Herrmann, Brach, Mertelsmann, and two colleagues—Albrecht Lindemann and Wolfgang Oster, who had defended their habilitations under Mertelsmann. Taking up where Gerok's investigation left off, Rapp's team examined the papers' figures. “We sent out questionnaires and asked the authors to disclose who was responsible for each,” says Rapp, whose group also requested raw data. He says they mostly received scant cooperation.

Still, the task force ferreted out cooked figures. They scanned every single autoradiogram and compared them with each other using image-processing software to alter orientation, intensity, and contrast. All told, Rapp says, his team found 357 apparent manipulations and fabrications of autoradiograms in the articles Herrmann co-authored. For instance, in one paper on an oncogene's activity, the same assemblage of bands was used to illustrate cellular responses to different chemicals. In other figures, bands were copied, flipped, or otherwise reused elsewhere. The 94 papers with clear or suspected manipulation could be “a lower estimate,” says Gerok, as the task force members were unable to check the veracity of the raw data in most papers. Herrmann, now in private medical practice, did not respond to several requests from the task force, nor did he return calls from Science. Brach could not be located.

The task force investigators also found evidence of fudged figures in the habilitations of Brach, Lindemann, and Oster. The latter two contained allegedly manipulated figures appearing exclusively in the habilitations. “Since a habilitation is a single-author publication, the author is ultimately responsible for its entire content, and this makes an active participation [in the manipulations] very likely,” says Rapp. During the initial investigations 3 years ago, Lindemann, who has denied involvement in data manipulation, retracted his habilitation and replaced it with a new version. Rapp's task force intended to examine it, but Lindemann failed to respond to a request for raw data, Rapp says. Instead, Mertelsmann informed the task force that Lindemann had decided to leave academic research. A University of Freiburg spokesperson, however, told Science that Lindemann is still with the university and has asked for a 1-year leave of absence starting 1 August. According to Rapp, Oster, who has since left the University of Freiburg, declined to respond to the task force other than with a written statement claiming he had nothing to do with the alleged misconduct. Neither Lindemann nor Oster could be reached for comment.

The task force report criticized Mertelsmann for failing to monitor the department's research closely enough to pick up on the fraud, “especially for publications in which he is the senior author,” says Gerok. During its investigation, Rapp says, Mertelsmann told the task force that as a clinician he lacked a deep understanding of the molecular biology research, and thus left it to Herrmann to oversee. That did not impress the panel—or Gerok, who says that “such a notion of authorship is lethal for science.” The task force also randomly selected five of Mertelsmann's 245 papers that do not include Herrmann as an author. After analyzing one in detail, the task force in its report states that it found “no conclusive proof of conscious falsifications,” although “data may have been manipulated.” Rapp's group plans to analyze the other four papers in the coming months. When contacted by Science, Mertelsmann declined to comment.

All the researchers implicated in data manipulation could face demands from the DFG and the cancer charity, the Deutsche Krebshilfe, to repay grant money used to fund tainted work. The DFG, Grunwald says, is considering legal action to recover part of the $440,000 it spent on Herrmann's work. In the meantime, Gerok says, the authors should retract all 52 clearly tainted papers. On 19 June, University of Freiburg president Wolfgang Jäger resurrected the university's own fraud panel in order to probe for irregularities in Mertelsmann's department. “You can expect this process to drag on for a long time,” says Rapp. Indeed, it may take years more for the German scientific community to put this affair behind it. 2. ASTRONOMY # Researchers Get Spectrum Bands 1. Gary Taubes Technology has a way of opening windows of opportunity and promptly closing them again. In radio astronomy, a surge of technological advances has opened such a window over the past decade by allowing astronomers to see the electromagnetic spectrum at millimeter wavelengths. But scientists have been in danger of losing this precious band to another burgeoning technology—telecommunications—as millimeter wavelengths look promising for transmitting high-bandwidth wireless information over relatively short distances. Earlier this month, however, astronomers won an international agreement to keep their window open. For astronomers, this band in the spectrum, which falls between long-wavelength radio waves and short-wavelength infrared light, is a gold mine. The bulk of the energy in the universe is emitted at these wavelengths; it comes from molecules in the interstellar medium, from the cosmic microwave background, and from cool material forming stars, planets, and galaxies. Studying these emissions promises insight into subjects as diverse as the origins of life and the birth of stars. But commercial technologies are already being developed to broadcast and receive at frequencies above 50 gigahertz, which corresponds to wavelengths below 6 millimeters. That's uncomfortably close to the frequencies of interest to radio astronomers, which fall between 71 and 275 GHz. “If you wanted, for instance, to broadcast the Internet from satellites in orbit—and God forbid this will ever happen—you have to go to much lower wavelengths than any employed now,” says astronomer Al Wooten of the University of Virginia and the National Radio Astronomy Observatory in Charlottesville. Now critical wavelengths have been guaranteed safe for radio astronomers. Although radiation in much of the millimeter band is blocked by Earth's atmosphere, three frequency bands can pass through relatively unscathed. The 2500 delegates to the World Radio Communication Conference, known as WRC-00, which met from 8 May to 2 June in Istanbul, Turkey, agreed to allocate all three to radio astronomers, reserving frequency bands with no scientific importance for military uses, such as satellites that require downlinks at millimeter frequency. “This is a preventive measure to assure that the future of millimeter-wave astronomy will be as free of interference as possible,” says Tomas Gergely, electromagnetic spectrum manager for the National Science Foundation's Division of Astronomical Sciences. When the WRC last dealt with spectrum allocations back in 1979, millimeter-wavelength astronomy was in its infancy. “At the time there was no way of emitting or detecting waves in that band,” says Wooten, “so there was no commercial interest, and the WRC decided to ignore it. Now, as the technology is developing, people are also discovering military and commercial applications in that band.” For radio astronomers, advances in receiver technology and high-speed digital processing have turned millimeter wavelengths into a new frontier. A$550 million joint European- U.S. observatory, to be built in the high desert of northern Chile, is at the end of its design phase. The project, which will consist of an array of 64 12-meter radio telescopes, is scheduled for completion in 2010. Some plan to use it to search the interstellar medium for complex molecules, such as glycoaldehyde, the simplest possible sugar, recently observed among the stars. Wooten describes such chemicals as “on the road to life.” Others will use the observatory to study galaxies billions of years older than the Milky Way that formed stars at a rate 10 to 100 times faster. “A lot of those galaxies are completely invisible at the visible wavelengths, because their photons have been redshifted all the way into millimeter wavelengths,” says Wooten. “Now we know we'll be able to see them clearly and without interference.”

3. NUCLEAR SECURITY

# Los Alamos Under Siege After Secrets Recovered

1. David Malakoff

Smaller than a paperback spy novel, the secrets-packed computer hard drives that temporarily disappeared at the Los Alamos National Laboratory in New Mexico could spark big changes in science and security at the country's nuclear weapons laboratories.

As Science went to press, investigators were trying to determine if the disks—missing from a vault since at least 7 May and discovered behind a nearby copying machine on 16 June—were pocketed by spies or just mislaid by employees. Their disappearance touched off a debate about how useful they might be to a rogue nation or terrorist group. Energy Secretary Bill Richardson says he believes the disks never left the premises and “espionage was not a factor.” But he isn't waiting for a final report to slap new controls on the flow of sensitive lab information. He has already blamed the lab's contractor, the University of California (UC), for the lax security, raising the possibility that the Department of Energy (DOE) may try to sever the university's 57-year oversight of the lab. At the same time, some lawmakers are calling for Richardson's head.

The incident has refocused attention on lab security and revived debate about the fate of Los Alamos scientist Wen Ho Lee, arrested last December and awaiting trial for allegedly mishandling classified information. It has also broken the logjam blocking the confirmation of former CIA official General John Gordon as head of a new National Nuclear Security Administration to improve security and oversee all weapons work. These and other issues were expected to get a high-profile airing at congressional hearings this week, even as several task forces and the FBI investigate how a team that is supposed to help prevent nuclear terrorism lost track of its classified cookbook for finding and disarming weapons.

To date, DOE officials have been intentionally vague about the contents of the laptop computer hard drives, confirming only that they stored information that might help its Nuclear Emergency Security Team (NEST) find, identify, and disarm a homemade atom bomb or stolen warhead. Formed in 1975, NEST has responded to dozens of calls with a team of scientists and emergency personnel equipped with sensitive bomb-finding and -disarming equipment.

Those familiar with NEST have speculated that the hard drives contain information, ranging from bomb radiation signatures to wiring diagrams, that could be valuable to terrorists and aspiring nuclear powers. Even poorly detailed guides to the shape and construction of weapons components, says Greg Mello of the nonprofit Los Alamos Project in Santa Fe, “would be very valuable to a technically advanced but data-starved country like Pakistan. It would shave years off new weapons' development by helping them avoid dead-end research alleys.”

A few commentators have proposed an even more frightening scenario: “The missing data also reveal how a stolen bomb might be set off,” Gary Milhollin of the Wisconsin Project on Nuclear Arms Control wrote on 16 June in The New York Times. Such fears led Senate Energy Committee chair Frank Murkowski (R-AK) to press DOE officials to confirm or deny that assertion at a hearing last week. They declined, citing security concerns.

Some specialists, however, doubt that even savvy terrorists would be able to defeat the multiple fail-safe devices that prevent an unauthorized user or an accident from detonating a weapon manufactured by one of the major nuclear powers. Although little is known publicly about Russian and Chinese weapons, U.S. and European warheads are known to carry “electronic combination locks,” called permissive action links (PALs), notes arms-control scholar Dan Caldwell of Pepperdine University in Malibu, California. PALs automatically disable a weapon if a user makes repeated guesses at the correct digital code, he says. Even a thief with the right code would still face formidable obstacles to detonating the weapon, as sensors must detect an exact sequence of pressure, acceleration, or temperature changes before triggering the conventional explosives that prime the nuclear reaction. In addition, the trigger mechanism is believed to be sealed in a tamper-proof barrier that disarms the weapon if it is pried apart or subjected to unusual electromagnetic bursts. For all these reasons, the idea that terrorists could use information on the NEST hard drives to arm a stolen weapon “doesn't seem to be the most plausible” scenario, says Mello.

Still, the disappearance of the disks has reignited a long-running debate in Congress over how to protect U.S. nuclear secrets. Richardson and several senators had long resisted appointing someone to lead the new nuclear security agency, saying that the organization undermines the secretary's authority and would hamper environmental cleanup and civilian science programs at the labs. That resistance evaporated just days after DOE revealed the loss of the hard drives, however, with the Senate voting 97-0 on 15 June to confirm Gordon.

More changes are on the way. In interviews after the disks were rediscovered, Richardson said he had already ordered the reintroduction of document tracking and other security measures abandoned in the early 1990s. He promised to penalize researchers involved in the disk episode, once identified, and he declared that officials at the University of California, which oversees both Los Alamos and California's Lawrence Livermore weapons lab, “have some explaining to do.” The university is “very strong on science,” he noted, but hasn't “done a good job” on security.

Six members of the House Commerce Committee, including Representative John Dingell (D-MI), want Richardson to dump UC. “It is time for [DOE] to take charge,” they wrote in a 16 June letter. Other critics, such as Milhollin, would like to give weapons work back to the Pentagon, which controlled it during and immediately after World War II, saying that it “has a much better security record.” UC's contract runs through 2002, however, and spokesperson Rick Malaspina says its “commitment to managing the labs remains strong.”

Meanwhile, many Los Alamos researchers are demoralized by the latest publicity and beg to be left alone. “Things were just getting back to normal after the fire,” says one scientist. “Now we're right back in the flames.”

4. PLANETARY SCIENCE

# Imaging Spat Pits Amateur Against Pros

1. Govert Schilling*
1. Govert Schilling is an astronomy writer in Utrecht, the Netherlands.

BostonA long-running dispute over who should get credit for first reporting landmarks on Mercury's uncharted hemisphere burst into public view on 26 May, when a Boston University press release claimed honors for a BU team without mentioning the contributions of an erstwhile collaborator, amateur astronomer Ron Dantowitz. The row, which has left both sides bitter and unwilling to work with each other, “was the opposite of how a collaboration between amateurs and professionals should be,” says one of the scientists involved.

Mercury is a major challenge to observe, because Earth-based telescopes must tease out the tiny planet's reflected light from the sun's glare. And Mercury orbits too close to the sun to be imaged by the Hubble Space Telescope, whose optics might be damaged by stray sunlight. Most of what we know about the scorched and pitted planet comes from images beamed back by Mariner 10, which in three flybys in 1974 and 1975 mapped half its surface. For the next quarter-century, the other half remained a mystery.

That was until Dantowitz, an education associate at the Boston Museum of Science, developed a technique called selective image reconstruction. It improves on the established “shift and add,” in which short-exposure photos, snapped in rapid succession, are aligned and combined for an overall sharper picture. The technique helps remove much of the blurriness caused by Earth's turbulent atmosphere. Dantowitz's improvements are faster imaging—using 1/60th-of-a-second frames from a high-speed video camera—and an old-fashioned eyeballing of each frame to select the sharpest parts of the best images, snapped during minimal turbulence, before creating the composite image. Testing his approach with light gathered by a 12-inch (30-cm) telescope at the museum's Gilliland Observatory, Dantowitz a couple of years ago captured sharp views of the Russian space station Mir and the U.S. space shuttle. Later, at the 60-inch (152-cm) Mount Wilson Observatory in California, he obtained high-resolution images of Jupiter and Saturn and their satellites.

In 1998 Dantowitz, with Scott Teare of the University of Illinois, Urbana-Champaign, was planning to observe Mercury from Mount Wilson when he was contacted by BU astronomer Michael Mendillo, who studies the planet's rarefied sodium atmosphere and wanted to cooperate on a project along with BU colleagues Jeffrey Baumgardner and Jody Wilson. “I offered to send the video feed of the high-speed camera to their digital recorder,” Dantowitz says. His group gathered data on the planet in late August 1998 and copied it, as promised, to a BU recording device.

Several months later, Dantowitz's team—including Teare and the museum's Marek Kozubal—finished its analysis, in which they described dark plains and at least one large, bright crater on Mercury's unmapped hemisphere. In June 1999, the trio submitted a paper to The Astronomical Journal.

The submission surprised the BU group, which learned about it through the Internet. They had been planning to list Dantowitz as a co-author on a paper featuring their own analyses. “We felt we had a collaboration,” says Baumgardner, “but he submitted his paper unilaterally without consulting us at all.” In a 24 August letter to Astronomical Journal editor Paul Hodge of the University of Washington, Seattle, Mendillo claimed that Dantowitz's paper was based on data belonging to the group and not Dantowitz's exclusive intellectual property. Hodge responded that he would hold up publication until the parties resolved the disagreement themselves.

After negotiations conducted mostly by e-mail—Mendillo and Dantowitz at the end were no longer on speaking terms—the two teams last February agreed to publish separate papers highlighting the newly observed mercurial features, which both appeared in last month's issue of The Astronomical Journal. The matter appeared settled, until BU issued a press release touting the technique and the images in advance of a presentation by Mendillo on 2 June at a meeting of the American Geophysical Union in Washington, D.C. Conspicuously absent was reference to the work of Dantowitz and his colleagues. In retrospect, says Baumgardner, “it probably would have been better if we had had a common press release.” With the bad blood, however, the teams may have Mercury's once-obscured face in better perspective than each other's point of view.

5. NEUROBIOLOGY

# Death Leads to Brain Neuron Birth

1. Elizabeth Norton Lasley*
1. Elizabeth Norton Lasley is a science writer in Woodbury, Connecticut.

Of all the body's organs, the brain seems least able to repair itself if damaged by injury, disease, or stroke. Indeed, throughout most of the 20th century, scientific wisdom held that neurons simply could not regrow after brain development ended. But in the past few years, scientists have provided mounting evidence of neurogenesis, or the production of new neurons, in some areas of the adult brain in organisms ranging from birds to mice and primates. One area that did not seem capable of such regeneration, however, was the neocortex—the region most concerned with such higher brain functions as memory and learning. But new work has now added the neocortex to the list.

In the 22 June issue of Nature, neuroscientists Sanjay Magavi, Blair Leavitt, and Jeffrey Macklis of Children's Hospital and Harvard Medical School in Boston report that when they induced certain neurons in the neocortex of adult mice to self-destruct, the loss triggered the formation of replacement neurons by brain stem cells. What's more, the newly formed neurons migrated to the same positions and made the same connections as their deceased predecessors. “This work shows that the adult brain has the capacity to respond to damage by repairing itself,” says neuroscientist Elizabeth Gould of Princeton University. If similar regeneration of brain neurons can be triggered in humans, the findings could open the door for treatments that might restore memory in Alzheimer's disease, for example, or undo the damage wreaked by spinal cord injury.

The current research is an outgrowth of previous findings in which Macklis and his colleagues showed that cell death, of all things, could foster a healing environment. Working with neurobiologists Constance Scharff and Fernando Nottebohm of Rockefeller University in New York City, Macklis had selectively induced apoptosis, a form of programmed cell death, in song-related areas of the brains of zebra finches. The result: a burst of neurogenesis. (The research was published in the 24 February issue of Neuron and was also described in Science, 25 February, p. 1381.) “But this was in a brain area and species where we know neurogenesis takes place,” notes Macklis. “The next question was, ‘Could we induce it where it does not normally occur?’”

To find out, Macklis and his Harvard colleagues zeroed in on a group of neurons in the mouse neocortex. Although new neurons were not known to grow in the area, it is near a potential source of neurons, because it lies above the subventricular zone, which contains so-called multipotential neural precursor cells, better known as stem cells. The researchers injected a select group of neurons in the neocortex with a light-activated chemical that triggers apoptosis. The resulting neuronal death mustered the underlying precursor cells to form new neocortical cells.

The researchers tracked the development of these neurons by giving the mice a tracer chemical called 5-bromodeoxyuridine (BrdU). BrdU is incorporated into newly synthesized DNA and thus labels cells that are dividing. The team then examined the BrdU-labeled cells in the neocortex for the presence of other markers that would indicate their developmental stage. These analyses revealed that the neocortices of animals given the apoptosis-inducing chemical contained new neurons in all developmental stages, ranging from those that were just born and were migrating up from the subventricular zone (detected by the presence of a protein called Doublecortin) to those that were fully mature (marked by the presence of the NeuN protein).

Other results indicated that the newly born neurons make functional connections. The damaged neocortical neurons originally sent their long axons into the thalamus of the brain. When the team injected the thalamus with a dye that's transported backward from the axon ends up to the nerve cell bodies, they found that the axons of the BrdU-labeled neurons picked up the dye. This finding suggests that the new neurons were making the same connections as those they had replaced. “We have evidence that by inducing apoptosis, we reactivated a program of developmental gene expression that was in place when the mice were embryos,” says Macklis.

Next, the researchers would like to identify the genes that control the neurons' development—a task that “will take the field many years,” Macklis cautions. But if it can be done, it might be possible to design drugs that reactivate the program in people who have suffered brain damage. Gould adds, “If we could figure out what the obstacles [to neuronal regeneration] are and how to overcome them, we might be able to get the brain to heal itself.”

6. CANADIAN UNIVERSITIES

# Funding of 2000 Slots Sets Off Musical Chairs

1. Wayne Kondro*
1. Wayne Kondro writes from Ottawa.

Edmonton, AlbertaJack Lightstone was thrilled last fall when the government announced a $605 million program to help Canadian universities attract and retain the best scientific talent (Science, 22 October 1999, p. 651). As provost and vice president of research at Concordia University in Montreal, Lightstone has spent the past 5 years building up the school's research capacity, and the government's promise to pay for 2000 new faculty positions across Canada over the next 3 years seemed like a godsend. This month the government fleshed out the details of its Research Chairs Program, awarding slots to 57 universities, including 21 to Concordia. But Lightstone is decidedly cooler these days toward the pending federal help. He's learned that at least six Concordia faculty members have already received “preoffers” of employment from other universities dangling the new chairs as bait. Those offers, say Lightstone and other university administrators, threaten to turn a program intended to stem a supposed brain drain to U.S. institutions into a game of musical chairs, forcing Lightstone and his colleagues to run faster simply to stay in place. Four months before it goes into effect, the chairs program has ignited a furor within Canadian academe. Benefiting from an allocation system based on a university's success in obtaining federal grants over the past 3 years, 15 large, research-intensive universities have received 70% of the 2000 slots. Leading the pack is the University of Toronto, whose 251 positions represent 8% of the total, whereas 28 of the 57 participating schools are getting fewer than 10 chairs, and about two dozen universities have been shut out entirely. The pot is supposed to be divided 45:35:20 among the natural, life, and social and behavioral sciences, although universities are free to decide the balance among individual disciplines and to create multidisciplinary posts. The chairs come in two sizes:$135,000 a year for established scientists, and $67,000 for rising stars. They are good for 5 years, with the senior awards renewable indefinitely and the junior slots good for a second term. Universities can use the money not just for salaries but also for travel, new equipment, and hiring students and postdoctoral fellows. In addition, the Canadian Foundation for Innovation (Science, 28 February 1997, p. 1256) announced last week that it will add$84,000 to the funds for each research chair to cover the institution's overhead costs.

In anticipation of the 1 September kickoff, the research-intensive universities have begun aggressively shopping for prospective candidates. Smaller universities say that has left them fending off talent raids. Short of a gentleman's agreement to eschew such raids, Lightstone says that the government should require all chairs to be advertised and filled through a competitive process. “Then you won't have one institution handing someone a chair on a silver platter, saying: ‘Come work for me.’”

Other administrators worry that the game of musical chairs will ratchet up overall costs by giving sought-after faculty members the leverage to push for higher salaries or better working conditions. “[In itself] that may not be bad,” says Michael Owen, vice president of research at Ryerson Polytechnic University in Toronto, which has received six chairs. “But some of the more senior and midcareer faculty members may feel somewhat slighted if they aren't seen as promising.” Rene Durocher, head of the chairs program, says such fears are unwarranted and that the program actually will give smaller universities the means to retain star researchers. “They are being alarmist,” he says. “If they lose some good people, they can now recruit some other people.”

Speaking with Durocher last month at a University of Alberta forum here sponsored by the Humanities and Social Sciences Federation of Canada, university administrators fretted that playing musical chairs will simply strengthen research-intensive universities and make it more difficult for have-not institutions to compete for research grants. They also see it fueling a trend toward a handful of megadepartments in certain disciplines.

Again, Durocher dismisses that concern. Although he admits that universities must make “tough decisions in what fields they want to develop, with whom,” he says that's preferable to having the government call the shots: “That would be micromanagement of universities.” The flexibility universities have to spend the money means that there will be “no losers,” he adds. And although established researchers may initially claim most of the positions, he predicts that ultimately the program will meet its intended goal of combating the brain drain by “helping Canada to retain its best researchers and to attract new superresearchers.”

7. GREENHOUSE WARMING

# Dueling Models: Future U.S. Climate Uncertain

1. Richard A. Kerr

When Congress started funding a global climate change research program in 1990, it wanted to know what all this talk about greenhouse warming would mean for United States voters. Ten years later, a U.S. national assessment, drawing on the best available climate model predictions, concludes that the United States will indeed warm, affecting everything from the western snowpacks that supply California with water to New England's fall foliage. But on a more detailed level, the assessment often draws a blank. Whether the cornfields of Kansas will be gripped by frequent, severe droughts, as one climate model has it, or blessed with more moisture than they now enjoy, as another predicts, the report can't say. As much as policy-makers would like to know exactly what's in store for Americans, the rudimentary state of regional climate science will not soon allow it, and the results of this 3-year effort brought the point home.

“This is the first time we've tried to take the physical [climate] system and see what effect it might have on ecosystems and socioeconomic systems,” says Thomas Karl, director of the National Oceanic and Atmospheric Administration's (NOAA's) National Climatic Data Center in Asheville, North Carolina, and a co-chair of the committee of experts that pulled together the assessment report “Climate Change Impacts on the United States” (available at http://www.nacc.usgcrp.gov/). “We don't say we know there's going to be catastrophic drought in Kansas,” he says. “What we do say is, ‘Here's the range of our uncertainties.’ This document should get people to think.” If anything is certain, Karl says, it's that “the past isn't going to be a very good guide to future climate.”

By chance, the assessment had a handy way to convey the range of uncertainty that regional modeling serves up. The report, which divides the country into eight regions, is based on a pair of state-of-the-art climate models—one from the Canadian Climate Center and one from the U.K. Hadley Center for Climate Research and Prediction—that couple a simulated atmosphere and ocean. The two models solved the problems of simplifying a complex world in different ways, leading to very different predicted U.S. climates. “In terms of temperature, the Canadian model is at the upper end of the warming by 2100” predicted by a range of models, says modeler Eric Barron of Pennsylvania State University, University Park, and a member of the assessment team. “The Hadley model is toward the lower end. The Canadian model is on the dry side, and the Hadley model is on the wet side. We're capturing a substantial portion of the range of simulations. We tried hard to convey that uncertainty.”

On a broad scale, the report can conclude: “Overall productivity of American agriculture will likely remain high, and is projected to increase throughout the 21st century,” although there will be winners and losers from place to place, and adapting agricultural practice to climate change will be key. Where the models are somewhat consistent, as in the far southwest, the report ventures what could be construed as predictions: “It is likely that some ecosystems, such as alpine ecosystems, will disappear entirely from the region,” or “Higher temperatures are likely to mean … a shorter season for winter activities, such as skiing.” Where the models clash, as on summer soil moisture over the eastern two-thirds of the lower 48 states, it explains the alternatives and suggests ways to adapt, such as switching crops.

The range of possible climate impacts laid out by the models “fairly reflects where we are in the science,” says Karl. But he notes that the effort did lack one important input: Congress mandated the assessment without funding it. “You get what you pay for,” says climatologist Kevin Trenberth of the National Center for Atmospheric Research in Boulder, Colorado. “A lot of it was done hastily.” Karl concedes that everyone involved would have liked to have had more funding delivered more reliably.

Even given more time and money, however, the assessment may not have come up with much better small-scale predictions, given the inherent limitations of the science. Even the best models today can say little that's reliable about climate change at the regional level, never mind at the scale of a congressional district. Their picture of future climate is fuzzy—they might lump together San Francisco and Los Angeles because the models have such coarse geographic resolution—and the realism of such meteorological phenomena as clouds and precipitation is compromised by the inevitable simplifications of simulating the world in a computer.

“For the most part, these sorts of models give a warming,” says modeler Filippo Giorgi, “but they tend to give very different predictions, especially at the regional level, and there's no way to say one should be believed over another.” Giorgi and his colleague Raquel Francisco of the Abdus Salam International Center for Theoretical Physics in Trieste, Italy, recently evaluated the uncertainties in five coupled climate models—including the two used in the national assessment—within 23 regions, the continental United States comprising roughly three regions. Giorgi concludes that as the scale of prediction shrinks, reliability drops until for small regions “the model data are not believable at all.”

Add in uncertainties external to the models, such as population and economic growth rates, says modeler Jerry D. Mahlman, director of NOAA's Geophysical Fluid Dynamics Laboratory in Princeton, New Jersey, and the details of future climate recede toward unintelligibility. Some people in Congress and the policy community had “almost silly expectations there would be enormously useful, small-scale specifics, if you just got the right model. But the right model doesn't exist,” says Mahlman.

Still, even though the national assessment does not offer the list of region-by-region impacts that Congress might have hoped for, it does show “where we are adaptable and where we are vulnerable,” says global change researcher Stephen Schneider of Stanford University. In 10 years, modelers say, they'll do better.

8. EUROPEAN SCIENCE POLICY

# France Rebels Against Gene-Patenting Law

1. Michael Balter

ParisFrance is on a collision course with the European Union over an E.U. directive that many researchers believe would allow raw DNA sequences of human genes to be patented. On 7 June, French justice minister Elisabeth Guigou told the National Assembly that the directive—which must be enacted by each of the 15 E.U. member nations by 30 July—contradicts French bioethics laws, which forbid the patenting of any part of the human body. If France maintains its defiance, it could be fined up to $600,000 daily for each day it refuses to adopt the directive. But the controversy may not be just another example of French pique: No E.U. country has yet adopted the directive, and two have mounted a legal challenge—although France is the only nation so far to hint that it might not toe the E.U. line if the legal maneuvers fail. Bolstering the French position is an Internet petition campaign against the law, which has gathered more than 4500 signatures, mostly from researchers and clinicians. The directive must be enacted by member states within 2 years of its adoption on 30 July 1998. One section would appear to ban the patenting of DNA sequences, stating that “the sequence or partial sequence of a gene” cannot be patented. But subsequent wording declares that “an isolated element of the human body … produced by a technical process,” including a gene sequence, can be patented “even if the structure of this element is identical to that of a natural element.” The latter provision, critics say, would seem to allow anyone to generate a human gene sequence and patent it without having to demonstrate its usefulness as an invention—say, as the basis for a drug. Axel Kahn, a geneticist at the Cochin Institute in Paris, argues that this wording could drive Europe into the same patenting frenzy that exists in the United States, where often tenuous claims of usefulness underlie efforts to try to put a lock on thousands of human genes (Science, 18 February, p. 1196). As an example of the potential consequences of the E.U. directive, an 8 June opinion by France's National Consultative Committee on Ethics backing the government's position cited a patent on the human gene coding for CCR5, a cell membrane receptor for certain immune signaling molecules. The patent was awarded in February to U.S.-based Human Genome Sciences Inc. Company officials have claimed that the patent covers use of CCR5, which serves as a coreceptor for the AIDS virus, to develop anti-HIV therapies even though the firm was unaware of the protein's role in HIV when it applied for the patent (Science, 25 February, p. 1375). That sort of sweeping claim appalls the directive's critics. “We absolutely must make a distinction between discovery [of a gene sequence] and invention [based on that sequence],” says French National Assembly deputy Jean-François Mattei, a pediatric geneticist who helped launch the petition drive last April (www.respublica.fr/sos.humangenome/index1.htm). Prominent French scientists such as Nobel laureates Jean Dausset and François Jacob have signed the petition, which calls for an “immediate moratorium” on the law's adoption to allow for its “renegotiation.” E.U. officials defend the law, which they argue is not at all ambiguous. A “technical procedure” that allows a sequence to be produced “outside the body, in a way nature cannot do itself … is patentable,” says Jonathan Todd, a spokesperson for the E.U.'s internal market directorate. Guigou's finding that the directive is incompatible with French law is irrelevant, Todd says: “[E.U.] law takes precedence over national law.” Also taking umbrage with opponents is France Biotech, an industry group that warned last week that failure to adopt the directive would “have disastrous effects” by leaving the genome field free to France's competitors, particularly those in the United States. Pascal Brandys, president of the French biotech company Genset, told Science that the distinction between a discovery and an invention insisted upon by critics “is not recognized in patent law. … This type of black-and-white definition is too simplistic.” The law's opponents are hoping France will convince its E.U. partners to rewrite the directive when it assumes the union's presidency on 1 July. They're heartened by the actions of Italy and the Netherlands, which are now challenging the directive before the European Court of Justice. But if these efforts bog down, says Todd, the European Commission may launch proceedings against France or any other E.U. member state that has not passed the law by 30 July. 9. RESEARCH POLICY # Report Tracks Federal Funding in Each State 1. Jeffrey Mervis Maryland is the most research-intensive state in the country, according to a new report* that describes in unprecedented detail where the federal government's annual$80 billion research budget is spent. The state's concentration of national laboratories and government contractors puts it second in total dollars received and far above California, Massachusetts, and other better known technology heavyweights in per capita federal R&D spending. That ranking is one of many facts in a 650-page tome by RAND's Science and Technology Policy Institute in Arlington, Virginia. Its authors hope the report will raise public awareness about research as well as inform politicians.

“It's hard to translate federal R&D into terms that members of Congress and the public can understand,” says lead author Donna Fossum, referring to the typical categorization by funding agency rather than recipient. “We decided that a state-by-state breakdown was the most manageable format. It also provides an answer for lawmakers who ask, ‘What's going on in my district?’”

View this table:

The report comes out just as several states, suddenly flush with budget surpluses, are pouring money into research as a way to stimulate their regional economies. “I think this information can be very useful for governors who want to leverage federal funding, or to fill in gaps,” says Tom Robel of the National Governors Association. Adds Dan Berglund, executive director of the State Science and Technology Institute in Columbus, Ohio, “it contains tremendously useful information that I've never seen before. In fact, we hope that the next Administration will decide to make this an annual report.”

• *Discovery and Innovation: Federal Research and Development Activities in the Fifty States, District of Columbia, and Puerto Rico.

10. NEUROSCIENCE

# A Critical Issue for the Brain

1. Marcia Barinaga

Do critical periods determine what we can learn and when? Neuroscientists and social scientists are probing the young brain to find the answer to this crucial question

Walk into a toy store, and you will see a vast array of “educational” aids for newborns and infants: flash cards to promote rudimentary math, videos to teach your baby to read. A colorful soft toy just doesn't seem to cut it anymore. Often, these products come with a not-so-subtle message: If you don't cram as much “learning” into your child's brain as possible before the age of three, she (or he) may never reach her full potential.

The notion that there is a “critical period” for learning in the first 3 years of life burst into the public consciousness after an April 1997 White House conference on early childhood development. The meeting drew on neurobiological evidence that a baby's brain is still developing after birth to reinforce the need for programs to ensure normal, healthy learning experiences for underprivileged children during all the years of childhood, including the first three.

“No neuroscientist ever got up there and said … that 0 to 3 was the most important time” for learning, says conference participant Carla Shatz, a developmental neurobiologist and chair of the neurobiology department at Harvard Medical School in Boston. But, says Shatz, a “break in logic occurred,” and in the ensuing publicity that message emerged. “As soon as this stuff hit Newsweek and every middle-class home in America had seen it, they totally lost sight of the fact that originally this was a motivational political message to address the needs of children who are really at risk,” says John Bruer, president of the James S. McDonnell Foundation, which funds neuroscience research.

Although the conference's message may have been misinterpreted—and misused—it has emphasized the need for a scientific understanding of the role played in learning and social development by so-called “critical periods.” These are defined as time windows when the brain is not only receptive to acquiring a certain kind of information, but indeed needs that information for its continued normal development. Critical periods have been well documented for the development of sensory systems in the brain, especially vision. But many neuroscientists also believe that critical periods exist for the development of at least some of the brain functions that underlie complex learning and thinking skills. There is “overwhelming evidence for critical periods,” for example in the learning of language, says neuroscientist Janet Werker, who studies the development of language at the University of British Columbia. But others caution that the concept is being used too broadly for too many types of learning, without rigorous demonstration that such windows exist.

“There are important critical periods in human development,” says William Greenough of the University of Illinois, Urbana-Champaign. “But there is remarkably little evidence to back up the notion that there are a lot of them.” Bruer—whose book The Myth of the First Three Years and other writings have established him as one of the most vocal skeptics about the role of critical periods in intellectual development—worries that the hype surrounding critical periods goes far beyond what the science supports and will push society to give up on children whose early years were disadvantaged.

One thing that the scientists agree on is that critical periods, where they do exist, are not as sharply defined as the popular message following the White House conference—or the pitches of educational toy manufacturers—suggest. No critical period ends suddenly, like a window slamming shut, but rather, they “taper off gradually,” says developmental vision researcher Terri Lewis of McMaster University in Hamilton, Ontario, who has studied critical periods in human visual development.

There is also no truth to the idea that critical periods are unique to the first 3 years. For those types of learning that seem to have them, the timing varies. And in many cases, the window never seems to shut completely, and the learning, albeit more difficult, can continue into adulthood. For that reason, researchers are coming to prefer the term “sensitive periods” to critical periods. The bottom line is that although it may be easier to learn a language or take up music as a child, adults can still do it. “The truth is somewhere in between [the two extreme views], as is often the case in science,” says pediatric neurologist Peter Huttenlocher of the University of Chicago.

## Early insights

There's no doubt that critical periods do exist for certain kinds of brain development. The most famous example comes from work done by David Hubel and Torsten Wiesel at Harvard in the 1960s. They and their co-workers showed that if a kitten's eye was closed for a time in infancy, the animal would be blind in that eye for life because the brain's visual system had missed out on the eye's input during a key stage of brain development. Other researchers, studying children who were born with crossed eyes or cataracts, have shown that critical periods also occur in human visual development.

Over recent decades, several types of studies have also suggested that the brain has sensitive periods for different types of learning. Some of the evidence comes from neuroscientists who have applied brain imaging and other techniques to study brain changes and correlate them with behavior and learning. Other evidence is entirely behavioral, deriving from psychiatric and education research.

Studies in this latter category have placed the most emphasis on the importance of the first 3 years. For example, psychiatric studies done in the 1950s found that children form an emotional attachment to their mother or primary caregiver during the first year or so of life. Numerous researchers have since shown that children who are securely attached, trusting their caregivers to protect and nurture them, form better relationships with others later in life than do children who do not enjoy such security. What's more, their research suggested that the bonds of attachment must be in place by age 3 for such benefits to occur. Developmental psychologist Ross Thompson of the University of Nebraska, Lincoln, who studies attachment, argues that during the course of human evolution, attachment would have been so key to an infant's survival that a critical period for its formation may be built into brain development.

The attachment studies were a major rationale for the focus on age 0 to 3 as a critically important time in a child's emotional development. But Thompson notes that several studies of children who spent their early years in Romanian orphanages, deprived of the type of normal human contact that provides an opportunity to form attachments, suggest “that the window of opportunity [for attachment] is far wider than we thought. Kids rescued at 4, 5, 6 can [still] form attachments.” But many of those attachments were weak or unhealthy, which may be a sign that the sensitive period tapers off at these later ages. However, without evidence of a clear end to the period, Thompson warns that it cannot be claimed as a bona fide sensitive period. An additional concern about drawing conclusions from the Romanian orphan studies, he says, is that these children were deprived in so many ways that there are many other potential reasons for their weak attachments.

Another social science study that suggests the importance of the first 3 years was conducted by education researchers Craig and Sharon Ramey of the University of Alabama, Birmingham, and Frances Campbell of the University of North Carolina, Chapel Hill. In an analysis of the impact of specially designed educational programs on more than 1000 children from underprivileged homes, they found that children enrolled from birth in preschool education programs showed greater improvement in IQ and school performance than did those enrolled in similar programs after school and during the summer in their early school years. That, says Craig Ramey, shows that opportunities missed in the first 3 to 5 years cannot necessarily be made up for later.

Studies such as these indicate that an enriched environment early in life is good for the brain. But Bruer argues that the results are being misapplied, especially by those who market “enrichment” products such as educational videos to middle-class parents. He notes that the Rameys and Campbell studied children from underprivileged environments. There is no scientific evidence, he claims, to suggest that additional enrichment beyond the normal environment for most children will offer any advantage.

Not only are middle-class parents being misled about what they need to provide their children, he says, but society's attention is being diverted from the needs of the truly underprivileged, which don't end at age 3. In support of this view, he cites an animal experiment that has been widely used to promote educational aids to enrich an infant's environment. Greenough, who performed the study, agrees with Bruer that his study has been misinterpreted.

In work done over the past 2 decades, Greenough and his colleagues have shown that rats raised in so-called “complex environments”—housed with other rats, and with lots of toys to play with—develop more of the neural connections called synapses in their brains than do rats raised alone in standard lab cages. But Greenough says his studies are “more of a deprivation experiment than an enrichment experiment,” because the complex environments are probably closer to the normal environment for a young rat than is the bleak environment of a typical lab-rat cage. His findings, he says, suggest that extreme deprivation is detrimental, but do not address whether additional stimulation beyond that found in a normal environment is necessarily better.

Greenough also says his study has little direct bearing on the case for enriched environments in a human infant's first 3 years, because his rats didn't enter the complex environments until they were weaned, which is equivalent to 2.5 to 5 years in human age, and they remained there until puberty. What's more, when the team exposed adult rats to similar environments, their neural connections proliferated also. “The changes occur faster in younger animals, and the magnitude is greater in younger animals, but [the effect] simply doesn't go away,” Greenough says. “It is very hard to read my work … and not to realize that it completely undermines [the] concept that everything is over by 3 years of age.”

## Windows for learning

Whereas the Ramey-Campbell work, as well as Greenough's, focuses on the brain's preparation for learning in general, other researchers have examined whether there are critical periods in the development of specific skills such as music and language. In a 1995 brain imaging study of musicians, for example, Thomas Elbert of the University of Konstanz in Germany and Edward Taub of the University of Alabama, Birmingham, found that the left hand of string musicians is represented by a larger area in the brain's touch-sensing region, the somatosensory cortex, than is the left hand of nonmusicians.

The researchers' main conclusion was that the brain's ability to change in response to music training extends into adulthood. But they also found that string musicians who began their study before age 12 had the largest brain area devoted to left-hand sensation, a possible indication that the brain is more receptive to musical training earlier in life. Elbert and Taub's data might be explained by a practice effect: Those who started young had simply been playing longer. Taub acknowledges that he and Elbert didn't systematically rule out that possibility, but he says the difference in total practice time between the two groups is too small to explain the magnitude of the difference in brain organization.

Researchers looking at the receptivity of the brain to language-learning have controlled for a practice effect, and they believe they have clear evidence that the young brain can learn a language with a fluency that is only rarely achieved by older learners. In other words, there appear to be sensitive periods for language acquisition.

Some of the strongest behavioral evidence comes from cognitive scientist Elissa Newport of the University of Rochester in New York and her colleagues. In the late 1980s, Newport, then at the University of Illinois, and graduate student Jacqueline Johnson studied 46 Chinese and Korean immigrants to the United States who became immersed in English at ages ranging from 3 to 39. To rule out a practice effect, the researchers matched their subjects for the number of years they had been using English. Then they played them recordings of spoken sentences, some of which had a grammatical error such as incorrect word order or the wrong verb tense, and asked them whether the sentence was correct.

For native speakers, the test is a breeze, says Newport, but in the immigrants the researchers found a “very systematic decline in correct responses as a function of the age at which people first arrived in the United States.” People who arrived before age 5 or so did as well as native speakers, Newport says, but “each group after that is systematically lower,” until the curve flattens out for those arriving after the teenage years. “It is exactly the shape you would expect from a critical period.”

One could argue, Newport says, that the later learners of English did poorly not because the critical period for language learning had passed, but because the greater number of years of experience with their first language simply interfered with the learning of another language. To address that issue, Newport and her husband, Ted Supalla, turned to a special group of people who did not learn any language as infants: deaf individuals whose hearing parents don't speak American Sign Language (ASL) and who learned ASL after entering a boarding school for the deaf at age 5 or 12. Newport and Supalla, who is himself deaf, compared a group of these delayed ASL learners to deaf ASL signers who had been immersed in ASL from birth. To rule out practice effects, they chose subjects who were 50 to 70 years old at the time of testing and had been using ASL for a minimum of 48 years.

The researchers tested production and comprehension of ASL sentences and “found the same kind of picture as in the second- language studies,” says Newport. Those who began to use ASL at age 5 scored on average slightly lower than those exposed since birth, while those who didn't begin until they were 12 scored lower still.

The results of behavioral studies on language development are buttressed by neurobiological findings. Neuroscientist Helen Neville of the University of Oregon, Eugene, has looked at brain organization in Chinese and Spanish immigrants who began to learn English at ages ranging from 2 to 16. Using brain imaging, she and her colleagues observed the subjects' brain activation patterns as they listened to sentences that had grammatical errors such as those used by Newport. “In people with delayed exposure to English, even as few as 4 years after birth, we already see a difference in the brain organization of the response to the grammatical surprise,” Neville says. In those who learned the second language before age 4, the response is entirely on the left side of the brain, where the language areas normally are, but later learners, says Neville, show more right-hemisphere activity. This suggests that the brain physically incorporates a late-learned language differently from one learned early.

Those results match what Newport and her colleagues saw when they looked at grammar ability. But grammar is only one element of language learning; other elements include phonology, the sounds of the language, and semantics, the meaning of words. And those don't necessarily have the same sensitive window. For example, Neville says, when she and her colleagues looked at the brain and behavioral responses to a so-called “semantic surprise,” a sentence in which one word doesn't make sense, “late learners look the same as early learners. It doesn't look as though there is a real tight critical or sensitive period.”

Even within one aspect of language, such as phonology, there may be different windows for learning. Some parts of phonology must be learned very early to master them like a native, while others can be learned well throughout life. That means, says Neville, that “language is not a single, monolithic, homogeneous system that either does or does not display a critical period.”

## The basis in the brain

The brain matures throughout childhood, and Neville and others suggest that steps in the maturation process may drive the timing of sensitive periods. For example, Chicago's Huttenlocher and his colleagues have counted synapses in the postmortem brains of children of various ages. They find that synapses proliferate in most brain areas during the first year of life, after which, he says, “you have a period when the synaptic density is high, for 6 to 12 months up to 5 to 15 years, depending on the area.” Then the synapse levels decline, with visual areas tending to lose their synapses first and the higher cognitive areas dropping to adult levels later. Harry Chugani at Wayne State University in Detroit and his colleagues have used positron emission tomography (PET) imaging to measure metabolic activity in the brains of infants and children, as an indirect way of looking at synapse production and elimination, and have come to similar conclusions.

Huttenlocher notes that the basic functions of a brain area emerge during the period when he sees that initial proliferation of synapses. For example, when the synapses begin to increase in the visual cortex, the child develops binocular vision. The pruning of synapses, Huttenlocher adds, appears to be associated with “the upper limit of easy learning of certain tasks,” at least to a first approximation. For example, while there seem to be different sensitive periods for different aspects of language learning, 12 to 14 years is roughly the age when the general ease of language learning declines, and, says Huttenlocher, “that is about the time during which the density and number of synapses in the language areas of the brain decrease.”

Despite correlations such as these, some neuroscientists and psychologists suggest that some apparent sensitive periods may be more a function of the cumulative nature of learning than of the physical development of the brain. Developmental psychologist Alison Gopnik of the University of California, Berkeley, has found that during their fourth year of life, children learn that other individuals have thoughts and views that differ from their own. Her work suggests that this happens when it does because the children have accumulated the experiences necessary to draw conclusions about the existence of other minds.

For example, she can accelerate the learning of these concepts by giving children special training that emphasizes the idea that others think differently. That suggests, she says, that there isn't “some maturational event in the brain” that makes the timing right, “but rather that the very things you learn enable you to learn new things.”

What it may come down to, at least for some types of complex learning, is a question of whether learning drives changes in the maturing brain, or whether the maturation process controls the ease with which learning occurs. Such questions can be addressed, says Neville, as brain structures associated with different kinds of learning are identified. For example, she is currently experimenting with children to see if training that accelerates their language learning results in measurable changes in brain organization, and several research teams are beginning to use brain imaging to investigate organizational changes in brain areas involved in the formation of bonds of attachment. “The work is going on; we just don't have the answers yet,” says Neville. But she predicts that any answers are “not going to be either-or. We have a whole panoply of brain systems. It is likely that the answer is going to be different for each individual system.”

As researchers pool their resources to nail down what role critical periods may play in learning, certain themes are emerging: Whereas younger brains may change more readily, older brains have not lost that capacity to change. And although it is clear that childhood is a privileged time for learning and one not to be wasted, there is no reason to give up hope for learning at any age. Indeed, says Rochester's Newport, the work may produce an understanding of whether the mechanisms of late-life learning differ from those of childhood. With a better understanding of such differences, says Newport, “one could think of different approaches and strategies” to improve adult education programs. And that would be good news for eager learners of all ages.

11. SOCIETY FOR DEVELOPMENTAL BIOLOGY MEETING

# A Mile-High View of Development

1. Gretchen Vogel

Boulder, ColoradoNearly 600 scientists gathered at the base of the Flatirons to discuss the growth and patterning of organisms including plants, worms, fruit flies, fish, and mice at the 59th annual meeting of the Society for Developmental Biology. Among the highlights were clues about how blind cave fish lost their eyes and how a gene that influences cell movement might help cancer spread.

A Fish's Tale “An eye for an eye and a tooth for a tooth,” declares the ancient biblical commandment. But for a population of blind cave fish, the exchange may have been an eye for a tooth—or several teeth. The theory is far from proven, but at the meeting researchers presented evidence that changes in the expression of a key gene involved in facial development might help explain the lost sight in cave fish. They suspect that during the course of evolution, the cave fish may have exchanged its sight—unneeded in underground rivers—for more teeth and taste buds.

The Mexican tetra fish (Astyanax mexicanus) thrives in habitats from surface waters to lightless caves in northeastern Mexico. Although technically the same species as their light-dwelling cousins and able to interbreed with them, the A. mexicanus from caves are much paler and have no eyes as adults. Intrigued by different-looking fish within the same species, developmental biologist William Jeffery and postdoctoral fellow Yoshiyuki Yamamoto of the University of Maryland, College Park, have been trying to understand how evolutionary changes in the animal's development caused the troglodytic fish to lose its eyes.

The researchers first examined the expression patterns of several eye-related genes in cave fish embryos, hoping to find clues about why the eye, which begins to develop almost normally, eventually degenerates and disappears as the fish matures. When they checked the young fish for the presence of one of the genes switched on early in eye development, called Pax6, they found a curious pattern: Throughout almost the entire cave fish embryo, the expression pattern of Pax6 matched that seen in surface-dwelling A. mexicanus. But in the region of the embryo destined to give rise to eventual eye cells, Pax6 was less prominent and appeared farther from the midline (a precursor of the backbone).

Previous work had shown that Sonic hedgehog (Shh), a fundamental patterning gene (named for a character from a children's computer game) that is active at the midline, can affect expression of Pax6. When Shh is missing, for example, the mutants develop so-called cyclopia—an enlarged single eye in the middle of the forehead. To see whether Shh might play a role in the loss of eyes, the scientists compared the pattern of Shh expression in cave fish and surface fish embryos. In cave fish, they found, Shh appeared in a wider swath at the midline, suggesting that the cave fish might have developed a sort of anticyclopia in which extra Shh protein causes smaller—or missing—eyes. “We were very surprised,” Jeffery says, “that our initial guess was actually correct.”

To see whether they could mimic the suspected evolutionary changes in the lab, the researchers injected excess Shh mRNA into surface fish embryos at the two- to four-cell stage. Indeed, the fish developed smaller-than-normal eyes. And when Yamamoto exposed early cave fish embryos to ethanol, which is known to interrupt Shh protein signaling, the cave fish seemed to develop slightly larger eyes. The scientists caution that experiments are preliminary—ethanol, especially, is likely having multiple effects on gene expression. Even so, evolutionary developmental biologist Brian Hall of Dalhousie University in Halifax, Nova Scotia, calls the findings “quite suggestive.”

Because the differences in gene expression appear so early, before the cells destined to become eyes have fully differentiated, Jeffery and Yamamoto suspect that the altered Shh pattern has caused other changes in cave fish facial development. To find out, the team plans to examine the effect of Shh on the development of teeth and taste buds. In a paper published this spring in Development, Genes and Evolution, Jeffery's team reported that cave fish have more taste buds than surface-dwelling fish have, and Jeffery says they also have more teeth, based on work with David Stock of the University of Colorado, Boulder. He suspects that during the course of evolution, more teeth and more taste buds may have given the ancestors of the bottom-dwelling cave fish an advantage. An ancestor of today's blind bottom-dwellers might have inherited an up-regulated signal from Sonic hedgehog, which resulted in more teeth but smaller eyes. Because vision was of little use in lightless caves, there was no evolutionary pressure to keep eyes.

That's an extremely interesting idea, says Hall. He cautions that one “can't make much of an evolutionary story out of one species” and suggests that additional research in other eyeless animals may be necessary. Even so, he says, the cave fish model “begins to provide a way of tackling how such coordinated evolutionary changes happen.”

Insights Into Cellular Travel

The intricately choreographed movements of cells during development have long posed fascinating riddles for biologists: How do cells know where to go, and how do they manage to get there? Now, researchers are learning how one group of specialized cells in the fruit fly moves during egg development—and in the process, they have uncovered new clues about how cancer cells might spread to form new tumors.

At the meeting, developmental biologist Denise Montell of Johns Hopkins University School of Medicine in Baltimore reported that she and her colleagues had identified a new gene needed for the movement of the so-called border cells in the reproductive tract of female fruit flies. During oocyte development, the six to 10 border cells migrate from the tip of the egg chamber to its center, where they eventually help form the shell that surrounds the mature egg. The gene that the researchers linked to these movements resembles one previously identified as involved in the spread of human breast cancer and may even work in a similar way: by cooperating with a steroid hormone, ecdysone in the case of the fruit fly, to regulate gene expression. If so, then the work may not only provide clues about how hormones influence development, but it may also help cancer biologists understand why exposure to hormones such as estrogen makes some cancers more aggressive.

Montell, with graduate student Jianwu Bai and postdoc Yoshi Uehara, found the gene by searching for mutations that cause border cells to move abnormally. Among the handful of mutants they identified was one in which the cells seemed to form properly, but migrated at a slower than normal pace. Bai named the mutant taiman, Chinese for “too slow.”

Thanks to the then partly finished fly genome project, the team was able to identify and clone the gene at fault in a matter of days. They found that it closely resembles a human gene called amplified in breast cancer-1 (AIB1), which takes its name from the fact that particularly aggressive tumors frequently contain extra copies of the gene, leading to overproduction of the AIB1 protein (Science, 15 August 1997, p. 965). Scientists think that excess AIB1 may help tumor cells spread, although they don't know exactly how. They do know that the protein works with activated estrogen receptors to help turn on genes in the nucleus, but have not yet identified the target genes. The evidence in flies suggests that taiman works in a similar way: When the team tested the interactions of taiman and the fruit fly's only known steroid hormone, ecdysone, in a cell culture assay, they found that adding more of the taiman gene increased the hormone's effects on gene expression.

Many cancer biologists had assumed that AIB1 and estrogen primarily promote growth, and that additional tumors were a side effect. But taiman's role in cell movement suggests that the human version might have a more direct effect on cell motility, Montell says. To find out more about how taiman influences cell movement, the team stained mutant border cells for e-cadherin, a protein required for border cell movements. The researchers thought that the slow-moving cells might be missing that protein, but found instead that they had higher than normal levels of e-cadherin on their leading edge.

That may not be completely surprising, as e-cadherin, which sticks out from the cell membrane and hooks onto cadherins on other cells, seems to play a dual role in movement. It can hold cells in place, but cells can also use it to pull themselves forward, which requires that they periodically break their links with one cadherin to move to the next. Montell and her colleagues suspect that taiman might be part of a pathway that breaks down cadherins, allowing a cell to let go of its current partner. “Clearly the taiman gene affects the invasive behavior of the border cells,” says developmental biologist Ruth Lehmann of New York University, although she cautions that the steps between taiman's cooperation with ecdysone, the reorganization of cadherin, and cell mobility are still unknown.

Another big question concerns whether AIB1 and estrogen might have similar effects on cadherin in cancer cells, enabling them to move throughout the body. In that case, Montell says, it might be useful to design antimetastasis drugs that work by blocking the effects of AIB1. Lehmann agrees. Indeed, she suggests, the fruit fly border cells might even be a useful assay for such drugs. If so, a gene named “too slow” may speed up progress in finding better ways to block cancer.

12. AMERICAN ASTRONOMICAL SOCIETY MEETING

# Cool Comets, Barren Clusters, and a Maxed-Out Universe

1. Robert Irion

Rochester, New YorkAbout 900 astronomers gathered 2 weeks ago near the birthplace of Eastman Kodak Co. to share their latest pictures of the sky at the American Astronomical Society's 196th Meeting. Notable findings pointed to a cold origin for comet Hale-Bopp, a nasty environment for extrasolar planets, and a maximum size for the biggest structures in the universe.

## Star Clusters Appear Hostile to Planets

Astronomers detect planets around other stars so frequently these days that each new sighting barely causes a ripple. Now, a research team has made a splash by spotting no planets at all. The Hubble Space Telescope struck out when it looked for planets within a swarm of a million suns, astronomers reported at the meeting. Although the search was tuned only to giant planets in tight orbits around their stars—the easiest kind to spot from afar—the result is still “startling and unsettling,” says planetary theorist Alan Boss of the Carnegie Institution of Washington in Washington, D.C. It may mean that clusters of stars are poorer habitats for planets than even pessimistic theorists had imagined.

Hubble's target was a globular cluster called 47 Tucanae, a dense knot of stars about 12,000 light-years from Earth. Astronomer Ronald Gilliland of the Space Telescope Science Institute in Baltimore, Maryland, and a team of two dozen planet hunters and theorists observed the cluster nearly continuously for 8.3 days last July, one of the largest chunks of time granted for a single Hubble project. They monitored more than 34,000 stars, searching for slight dips in their brightnesses caused by giant planets crossing in front. Astronomers used that approach last year to spy a planet whizzing around a nearby star once every 3.5 days (Science, 19 November 1999, p. 1451). About 1% of stars near our sun appear to have these so-called “hot Jupiters” in tightly bound orbits.

If that same statistic applied to 47 Tucanae, Gilliland and his colleagues should have seen distinctive dimming patterns from 15 or 20 planets. Instead, they saw none. Hubble's 1300 exposures of the cluster were sensitive enough to reveal planetary transits with ease, says astronomer Tim Brown of the National Center for Atmospheric Research in Boulder, Colorado. “We're convinced that if there are real planets there, we should see them,” he says. “Something is clearly different about the cluster environment.”

Modelers are having a field day trying to explain why the cluster is so barren. “It's a theorist's dream,” says astrophysicist Steinn Sigurdsson of Pennsylvania State University, University Park. “Nothing has been ruled out.” A leading culprit, says Boss, is the cluster's composition. Compared with our sun and nearby stars, globular clusters contain far lower concentrations of elements heavier than hydrogen and helium. The ratio of iron to hydrogen in 47 Tucanae, for instance, is one-fifth that in our sun. Most other clusters have even less iron, carbon, oxygen, and other planet-building substances. Without enough grains of dust and ice to clump together and get planets started, the gas swirling around most young stars in a cluster may never have coalesced into Jupiter-sized objects.

If any planets did arise, the cluster's stars are packed so close together—about a million times more densely than in our galactic neighborhood—that passing stars might wreak gravitational havoc and fling planets from their orbits. Indeed, astronomer Rex Saffer of Villanova University in Pennsylvania reported tantalizing clues at the meeting that stars may have collided and merged in the center of another globular cluster, called NGC 6397. “A stellar flyby can disrupt a planetary system, and physical collisions can occur with amazing frequency,” Saffer says. Gilliland agrees that close encounters might spell doom for planets, especially near the cores of clusters.

Still, Sigurdsson notes that hot Jupiters might survive gravitational perturbations because of their compact orbits. With astrophysicist Douglas Lin of the University of California, Santa Cruz, Sigurdsson is exploring yet another way to destroy such planets. Passing stars may jostle a gas giant enough to make its orbit slightly more eccentric. This will subject the planet to repeated bouts of tidally induced heating from its parent star, like the tugs endured by Jupiter's inner moon Io. Too many such nudges may overheat the planet, swelling it to the point of bursting.

No matter which scenario holds up, Hubble's results suggest that researchers with the Search for Extraterrestrial Intelligence project were misguided when they beamed a radio greeting toward stars in the giant globular cluster M13 in 1974. “Things are pretty hectic in a globular cluster,” Boss says. “If Hubble couldn't see any hot Jupiters, there might not be any planets at all.”

## Galaxy Survey Surpasses the Biggest Clusters

When your car's odometer rolls past 100,000 miles, you might buy a new car. Or you might just cheer and keep driving. Astronomers are choosing the latter option to applaud an Australian-British project called the 2dF Galaxy Redshift Survey, the first survey to gauge the motions in space of more than 100,000 galaxies. That figure—four times the number of galaxies charted in any previous survey—is more than an arbitrary milestone. It also marks the first time that astronomers have clearly extended their vision far beyond the largest coherent groups of galaxies that exist, known as superclusters. “We've seen the end of the biggest structures in the universe,” says team member Karl Glazebrook of Johns Hopkins University in Baltimore, Maryland. And as the survey pushes toward its goal of 250,000 galaxies, it is sharpening estimates of the amount of mass in the universe.

The survey, named for the Two-Degree Field spectrograph at the 3.9-meter Anglo-Australian Telescope in New South Wales, Australia, sees hundreds of superclusters with typical sizes of 200 million light-years across. The first such structure, the famed “Great Wall,” surfaced in the 1980s during a survey of our cosmic neighborhood by astronomers at the Harvard-Smithsonian Center for Astrophysics (CfA) in Cambridge, Massachusetts. Since then, more extensive surveys have seen filaments of similar size, notably the Las Campanas Redshift Survey at the Carnegie Observatories in Chile. That survey of 25,000 galaxies, completed in 1994, suggested that superclusters did not grow larger on ever-grander scales. “The 2dF maps confirm and strengthen that qualitative conclusion,” says astrophysicist David Weinberg of Ohio State University in Columbus.

Other researchers are pleased to see 2dF cross the six-digit galaxy threshold. “For a long time, this will be the standard in the field,” says CfA astronomer John Huchra, who with colleague Margaret Geller detected the first Great Wall. “2dF is a major step forward,” agrees astronomer David Helfand of Columbia University in New York City. Because the survey extends about five times more broadly and deeply into the universe, he says, “we now have a much more confident understanding of what the large-scale structures are like.”

At the meeting, 2dF researchers drew upon their observations to announce an independent new measure of how much matter the universe contains. The estimate relies upon the degree to which superclusters are “flattened” by the gravitational attraction of all of the matter within, including dark matter that telescopes cannot see. That flattening is evident as distortions in the redshifts, which measure how fast galaxies recede from us as the universe expands. Superclusters exert a pull toward their centers, says astronomer Gavin Dalton of the University of Oxford. That causes galaxies on the near sides of the structures to move away from us faster than they otherwise would, while galaxies on the far sides are slightly impeded.

If superclusters were extremely massive, Dalton says, this effect would squash the redshifts of galaxies around them much more than the 2dF survey reveals. To date, the team's analysis suggests that the total mass associated with galaxies accounts for just one-third of the “critical density” needed for the gravitational pull of matter to halt the expansion of space. That's consistent with recent studies of the light from distant supernova explosions and the faint glow of microwaves from the early universe that bathes the entire sky (Science, 28 April, p. 595). The mass estimates will improve as the 2dF group collects more galaxies—the 250,000 goal should be reached by the end of 2001, Dalton notes, with a robotic spectrograph that can capture up to 400 redshifts per hour.

Astronomers with a competing project now under way, the Sloan Digital Sky Survey, hope to surpass 2dF by tracking the redshifts of up to 1 million galaxies within 5 years. Dalton welcomes the chance for the teams to compare their statistics on galaxy clustering. “If you're going to measure something this complicated, it's a good idea to measure it twice,” he says. Still, Helfand thinks 2dF might have played the trump card: “I suspect 2dF is going to tell us everything about large-scale structures that Sloan will tell us.”

## Rare Gas Pinpoints Hale-Bopp's Cradle

Comet Hale-Bopp, which lit the sky gloriously 3 years ago, apparently was born in a frigid region of the outer solar system. Hale-Bopp journeyed toward the sun from a vast reservoir of comets called the Oort Cloud. Trillions of dark bodies may drift within this sphere, which extends at least a light-year into space. Comets didn't begin their lives there, however; astronomers think they arose among the gas-giant planets when the solar system coalesced. Gravitational kicks then expelled comets to the Oort Cloud, where they preserve traces of the solar system's early composition. Now, the first faint whiff of one primordial gas from a comet suggests that Hale-Bopp formed near Neptune, farther from the warmth of the sun than some astronomers expected.

The gas is argon, a “noble gas” with the same inert chemistry as helium, neon, and krypton. These substances don't interact with other elements or compounds, so they remain unaltered within a comet's icy matrix for billions of years unless they diffuse out of the nucleus. They boil into space if temperatures get too high—more than about 25 kelvin (-248°C) for neon and 40 kelvin for argon. This property makes the gases sensitive indicators of a comet's temperature history. But noble gases are notoriously difficult to detect in space, so astronomers had tried in vain for years to spot one in a comet's ethereal glow.

It took the bright spectacle of Hale-Bopp for the signature of argon to become “weak but unmistakable,” planetary scientist Alan Stern of the Southwest Research Institute (SwRI) in Boulder, Colorado, reported at the meeting. Others are delighted by the discovery, although interpreting the noisy data still leaves some room for doubt. “If the argon detection is valid, it would be very significant for constraining where the comet arose,” says astronomer Michael Mumma of NASA's Goddard Space Flight Center in Greenbelt, Maryland.

Stern and a half-dozen collaborators launched a sounding rocket 300 kilometers above the White Sands Missile Range in New Mexico to study Hale-Bopp on 29 March 1997, within a day of its closest approach to the sun. The rocket carried a small telescope and a spectrograph that Stern aimed at the comet for just 5 minutes, using a joystick and radio telemetry. In the complex spectrum of sunlight reflected by the comet, the team found a pair of faint wiggles at ultraviolet wavelengths characteristic of argon. “Hale-Bopp was such a boomer that there was just enough argon for us to detect,” Stern says.

The implication, he maintains, is that the comet arose between the orbits of Uranus and Neptune, where temperatures in the embryonic solar system hovered around 40 kelvin. That places tighter limits on Hale-Bopp's cradle than research by Mumma and others in 1997 with the Extreme Ultraviolet Explorer satellite, which found that neon was depleted in Hale-Bopp by a factor of at least 25 compared with the sun. If all of Hale-Bopp's neon burned off, its birthplace must have been warmer than 25 kelvin (Science, 5 September 1997, p. 1488).

Recent models by SwRI planetary scientist Harold Levison and colleagues suggest that Jupiter, Saturn, Uranus, and Neptune each may have injected comparable numbers of comets to the Oort Cloud. However, the solar system's disk of gas and dust was so much warmer near Jupiter that all of Hale-Bopp's argon would have wafted into space if it spent much time there, Stern says. “This comet must have been in a deep freeze,” he says, until some slight perturbation—perhaps from a passing star—nudged it inward from the Oort Cloud relatively recently.

That scenario is feasible, says planetary scientist Jonathan Lunine of the University of Arizona, Tucson. However, he warns that comets may have undergone chemical processing after their birth. Such modifications could have changed the original gas contents of comets in ways that astronomers still don't understand. “We cannot yet say that the comet formed at 20 or 30 kelvin just because we see argon,” he says. Stern notes that even if such changes occurred, the presence of argon shows that the comet's interior never rose above 40 kelvin, pointing to a quick exit from the solar system.

Astronomer Lucy McFadden of the University of Maryland, College Park, and others had hoped to see argon from Halley's Comet in 1986 with a shuttle mission, but the Challenger explosion foiled those plans. Stern's team produced “a very high return for a cost-effective observation,” she says. “These are remarkable results.”

13. PALEONTOLOGY

# Feathers, or Flight of Fancy?

1. Erik Stokstad

A controversial paper aims to turn avian origins on its head. But mainstream paleontologists say “put up or shut up”

Longisquama insignis never asked for all this fuss. Some 220 million years ago, the squat, mouse-sized reptile eked out an unassuming existence in what is now central Asia, gobbling Triassic insects and minding its business. But Longisquama was marked by destiny: Sprouting from its spine were at least six vanelike appendages up to 12 centimeters long—features unlike anything known to have graced a reptile before or since. Scientists disagree whether the creature used its plumes for gliding; they may have been a sexual display. Now, after decades in a Russian research institute, the appendages have propelled Longisquama's fossilized remains from relative obscurity and thrust it dead center into a bitter debate.

On page 2202, a team led by John Ruben of Oregon State University in Corvallis describes Longisquama's curious appendages in the greatest detail yet. In a radical interpretation of the fossil evidence, the authors argue that the appendages are feathers much like those of modern birds. Outside the paper, Ruben and his iconoclastic band go much further, touting the fossil as “an ideal bird ancestor.” That conclusion has infuriated paleontologists—not just because it challenges the prevalent theory that birds evolved from theropod dinosaurs, but because they say it does so in an unscientific way.

The minority of scientists who reject the dinosaurian origin of birds are elated by the new description of Longisquama. “It's almost too good to be true,” says Storrs Olson, curator of birds at the Smithsonian Institution, who did not contribute to the paper. “This is a bigger step forward to understanding the origin of birds than Archaeopteryx“—the 145-million-year-old fossil bird whose teeth, scales, and other primitive features forged an unequivocal link between birds and reptiles. The dinosaur advocates' response to this new nontheropod ancestor to birds? “Nonsense.” “Nuts.” “Rubbish”—not to mention several unprintable comments.

The fossil causing the flap was discovered decades ago in an ancient lake bed in what is now Kyrgyzstan. The Russian paleontologist Alexander Sharov published the first description in 1970. Since then, the specimen has remained at the Russian Academy of Sciences' Paleontological Institute in Moscow, where few Western paleontologists have had a chance to examine it. Ruben and his graduate student Terry Jones first saw it in early 1999, when Longisquama was touring the United States as part of a privately sponsored fossil show. After the tour ended, Russian paleontologists brought the fossil to Larry Martin's lab at the University of Kansas, Lawrence, where they examined it with Ruben, Martin, Alan Feduccia of the University of North Carolina, Chapel Hill, and others.

What riveted their attention was the creature's plumage. Previous workers had described the long, thin appendages as scales (the name Longisquama means “long scale”), but to Ruben and colleagues they bore a breathtaking resemblance to bird feathers. For example, they note that Longisquama's appendages have a central shaft with narrow ribs that extend out to the edges, roughly comparable to barbs on a feather. And near the spine, the shafts have wide, tubular bases, similar to the hollow calamus of modern feathers. The calamus grows from the circular epidermal ring around a follicle in a developmental pattern unique to feathers—a pattern Ruben thinks that the base of the shafts in Longisquama shares. “It had to have developed in a follicle,” he says. “There's no other way to do it.”

In short, Ruben and colleagues conclude, Longisquama was a feathered creature. That's important, Ruben says, because he and most other scientists think structures as specialized as feathers must have evolved only once, almost certainly in some ancient member of a vast group of creatures called archosaurs. Archosaurs include birds, dinosaurs, pterosaurs, crocodiles, and less well known animals, all of which share a unique pattern of openings in the skull.

But if only one branch of the archosaur family developed feathers, which one was it? Theropod dinosaurs, most paleontologists say, citing numerous skeletal similarities that link them to birds. Ruben disagrees. If a nondinosaurian archosaur—Longisquama—had feathers, he says, then it's difficult for him to picture dinosaurs' evolving them independently and turning into birds. On the other hand, “I can easily imagine avian evolution going through a Longisquama-like stage,” Ruben says.

When paleontologists hear statements like that, they reach for their entrenching tools. The problem, they say, is not that the authors of the Longisquama paper are presenting a rival hypothesis about bird origins; it's that they aren't really stating a hypothesis at all. Simply arguing that Longisquama is birdlike is not enough, says paleontologist Kevin Padian of the University of California, Berkeley. You must be able to show that it is more closely related to birds than something else, such as theropod dinosaurs, he says.

To make such comparative evolutionary arguments, mainstream biologists and paleontologists in North America use computer-generated family trees called cladograms. First they choose anatomical characters, such as the length of digits or the presence of openings in the skull. Then they tabulate which animals do or don't have them, and run the resulting matrices through computer programs that rank the possible evolutionary lineages in order of simplicity. The most parsimonious cladogram is the one that requires the fewest evolutionary changes. Other things being equal, that family tree, or phylogeny, is the one that scientists prefer. “This gives us an explicit, repeatable, falsifiable basis for conducting our studies,” says John Merck, a paleontologist at the University of Maryland, College Park.

But to the chagrin of their opponents, Ruben and others who object to the dinosaurian origin of birds don't argue that way. Instead of postulating cladistic relationships, Merck charges, Ruben and others present fossil evidence that they say suggests something other than dinosaurs might have given rise to birds, but without fleshing out alternative scenarios that others can test. “They're not playing the game,” says Tom Holtz, a theropod expert at the University of Maryland, College Park. “Just saying ‘This might have something to do with bird origins’ doesn't give us something that's falsifiable.”

For Padian, Longisquama is the last straw. “If you think this is where birds came from, where is your cladistic analysis?” he demands. “Where's your falsification of the dozens of cladograms that put birds right in the middle of theropod dinosaurs?”

Unnecessary, Martin says. “We don't have to do a cladogram. We can tell you right now that all the characters we found on the specimen were consistent with it being related to birds.” As for the many cladograms that demonstrate a dinosaurian origin of birds, Martin charges that they are riddled with characters based on mistaken anatomy—in other words, “garbage in, garbage out” on a massive scale.

But anatomy may prove to be the Achilles' heel of the Longisquama paper. The evidence for feathers is marginal at best, says Longi-skeptic Rick Prum, curator of ornithology at the University of Kansas's Museum of Natural History and one of the scientists who saw the fossil while it was in Kansas. “I think they've documented through careful observation that the bases [of the appendages] are cylindrical,” Prum concedes. Still, that doesn't mean they grew like feathers, he says, stressing that “everything else about these structures has nothing to do with feathers.” What they really look like, Prum believes, is ribbed membranes. Feathers tend to fray at the edges, he notes; Longisquama's plumes are fused. Near the base, where the appendage emerges from the skin of the beast, the ribs extend from the shaft backward toward the body, rather than toward the tip as a feather barb does. In one place, where two plumes cross, the one on top shows the clear imprint of the one underneath—something a feather wouldn't do. “The more you look at these things,” Prum says, “the less they are like feathers.”

Some paleontologists even doubt whether Longisquama is an archosaur. Hans-Dieter Sues, an archosaur expert at the Royal Ontario Museum in Toronto who has studied the specimen firsthand, points out that cracks in the skull make it impossible to find two key diagnostic features: an opening in the lower jaw and an opening in front of the eye socket. Sharov included them in a sketch accompanying his original description of Longisquama, but others who have examined the fossil have been far less confident that they are there.

But even paleontologists who dismiss any connection between Longisquama and the origin of birds still find it an intriguing fossil. “Wherever Longisquama fits on the tree of life, it's a remarkable reptile from an extremely interesting time in vertebrate history,” Holtz says. Prum points out that no modern animals have such long, membranous scales. Even if they have nothing to do with feathers, he says, “these structures are quite bizarre and fascinating. I want to know what they are.”

14. # Ground Zero: AIDS Research in Africa

1. Jon Cohen*
1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl.

Scientists have stepped up efforts against HIV at the epicenter of the epidemic, as the virus continues to exact a growing toll in human lives

On a clear and cheery rain-washed March morning in Kampala, Uganda, Rose Busingye, an AIDS social worker and nurse, leads four visiting Italian AIDS researchers on a dark and depressing tour that begins at an orphanage. “They have no one,” Busingye explains as she walks through the well-kept house that serves as the new home for a dozen children whose parents recently died from AIDS. She introduces a boy whose mother, a prostitute, hanged herself when she learned that her blood test for HIV was positive. A girl walks by whom Busingye had found wrapped in her dead mother's arms. Busingye points to siblings the orphanage found eating from a garbage pile.

The tour proceeds to Namuongo, a nearby village, where Meeting Point, the group Busingye works with, provides care and support for several people near death from AIDS. The researchers walk through fields of sugarcane, jackfruit, and yams, arriving at a cinderblock home with a corrugated tin roof. Inside, a 36-year-old woman, who has tuberculosis and the disorientation typical of AIDS dementia, is bedridden in a dark, two-room space she shares with her brother, his wife, and four children. She has been in this bed, a thin mattress on the floor, for 3 weeks now. “I'm not OK,” says the woman, who has a large red Bible by her side.

The last stop on the 3-hour tour is Kireka, a poor village that abuts a rock quarry. Packs of children roam the small slum, which has rows of huts made from mud, sticks, and straw. Busingye says 40 of these children have lost both parents—either because of AIDS or the civil wars here and in neighboring Rwanda—and now are cared for by extended family. One orphan with withered limbs, who himself has late-stage AIDS and is wearing nothing but a long-sleeved pink turtleneck, takes the hand of an Italian AIDS researcher and walks with her. Meeting Point helps 120 AIDS patients in Kireka, including a 32-year-old mother of two who lies on the ground in the village's marketplace selling vegetables. Her husband died from AIDS in December 1998. “He had another woman,” she explains. Now she has AIDS, as does her 15-year-old son.

Mario Clerici, a leading AIDS immunologist from the University of Milan, purses his lips and shakes his head. “All AIDS researchers should come here to see what AIDS is,” Clerici says. “Most AIDS researchers have barely seen AIDS patients. Here, you see how horrible it is.” Clerici holds up a hand and rubs his fingers together. “Come here and you can touch it.”

More and more European and North American AIDS researchers are now coming to Uganda and other countries in sub-Saharan Africa, which is home to an estimated 24.5 million HIV-infected people—a whopping 70% of the global total, even though this region has a mere 10% of the world's population. Like Clerici, these investigators are not simply coming to see the epidemic at its center, but to collaborate with local researchers on projects that aim to slow both HIV's spread and the course of disease in the millions already infected. “There really is an expansion, I'd even say an explosion, over the last few years of AIDS projects in Africa,” says Peter Piot, an epidemiologist who did pioneering AIDS research on the continent in the early 1980s and is now director of UNAIDS, the Joint United Nations Programme on HIV/AIDS.

But even with this explosion of research, the amount spent on African studies is a fraction of the total devoted to efforts in developed countries. Most cash-strapped African countries also spend little on research themselves, relying instead on the generosity of foreign collaborators, which inevitably raises tricky questions about research direction, control, and credit. More troubling still, most African countries—constrained by limited resources, weak infrastructures, social mores, and political inaction—have grave difficulties translating research insights into prevention and treatment strategies. “It's slower than anyone would like,” says Dorothy Mbori-Ngacha, a pediatrician at the University of Nairobi who studies ways to thwart the transmission of HIV from mother to child. “You can do very good work, and it can end up not changing anything if you can't package it in a way that policy-makers can use it. We have quite a lot of research sitting on shelves.”

From 9 to 14 July, the world's attention will focus on these and other HIV dilemmas faced by African countries, as thousands of researchers gather in Durban, South Africa, for their biannual international conference. Many Africans have high hopes that the XIII International AIDS Conference, the first ever held on this continent—or in any developing country—will live up to its motto and “break the silence” that, as chair Hoosen Coovadia put it in a welcoming note to delegates, “prevails across nations estranged by colossal inequities.”

## The modern plague

The surge of interest in African AIDS research mirrors the surge in HIV itself. As the World Bank noted in an unusually impassioned June 1999 report, “Intensifying Action Against HIV/AIDS in Africa,” only one African country had an HIV prevalence rate in adults higher than 2% in 1982. By 1998, UNAIDS estimates, HIV had infected more than 7% of the adults living in 21 African countries, and in two populations a staggering 25% were infected (see maps). “Intensifying Action” also stresses that the African death toll from HIV—which is transmitted by heterosexual sex in 90% of the cases—soon will eclipse the 20 million Europeans killed during the 14th century plague.

Although hot debates continue to surround the question of how and when HIV entered humans and began its spread around the world, several compelling studies suggest it is a variant of viruses that infect nonhuman primates in central Africa (see p. 2164). The oldest HIV sample, which comes from the Democratic Republic of Congo (formerly Zaire), dates back to 1959, but most researchers suspect that AIDS had been killing humans for many decades—if not centuries—before that.

UNAIDS has the best documented estimates of the current prevalence of HIV and AIDS, but even these numbers are squishy. Bernhard Schwartländer, the UNAIDS epidemiologist who heads the department that provides the estimates, notes that researchers derive the numbers by extrapolating from “sentinel” populations, such as pregnant women who come to clinics. “We're more and more confident of our estimates we're giving year after year, but there's an order of magnitude of uncertainty that we have to deal with,” says Schwartländer. Indeed, the HIV/AIDS estimates that UNAIDS will release later this week will have plus-or-minus values for the first time. (The global total living with HIV, for example, is 34.3 million ± 25%.)

The figures portend a grim future: Because of the long time lag between infection and disease, sub-Saharan Africa has only just begun to suffer the full brunt of the devastation AIDS ultimately will cause. “The most disturbing long-term feature of the HIV/AIDS epidemic is its impact on life expectancy, making HIV an unprecedented catastrophe in the world's history,” “Intensifying Action” concludes. “HIV/AIDS has already reversed 30 years of hard-won social progress in some countries. Now is the time for Africa—and the world—to fight back.”

## Fighting back

Finally, the World Bank, governments, industry, and nonprofit organizations have woken up to this unfolding disaster and are stepping up the fight against HIV/AIDS in Africa. “We have a fantastic opportunity now,” says Michel Pletschette, a European Commission official who oversees funding of AIDS research in Africa. “There's more will to do more. But it's a little bit sad to see we had to come to a situation where AIDS has a mortality equal to plague in the Middle Ages. If all these people were dying in a week, you'd see things differently than when they die in 10 years.”

Attention to Africa's AIDS problems gained momentum in May 1999 when the South African government, in collaboration with the electrical company Eskom, pledged $8 million to support the newly formed South African AIDS Vaccine Initiative, which will fund R&D. Later that month, Bristol-Myers Squibb, maker of two anti-HIV drugs, launched “Secure the Future,” a$100 million program to help five countries in southern Africa combat AIDS; it has already helped build a $5 million lab in Botswana to study drug resistance. A few weeks later, the Bill and Melinda Gates Foundation made a$25 million donation to the International AIDS Vaccine Initiative, which now is supporting R&D on four vaccines tailor-made for Uganda, Kenya, and South Africa. And international media are also paying attention: In November and December, the Village Voice, a free weekly paper based in New York City, ran reporter Mark Schoofs's remarkable eight-part series on AIDS in Africa, which later won the Pulitzer Prize.

This year's U.S. budget includes a $100 million program, the Leadership and Investment in Fighting an Epidemic (LIFE) initiative, that aims to slow the transmission of HIV in sub-Saharan Africa, improve treatment for infected people, and provide care for AIDS orphans. In a January speech during the United Nations (U.N.) Security Council session on AIDS in Africa, Vice President Al Gore declared that “we are putting the AIDS crisis at the top of the world's security agenda” and pledged that the Clinton Administration would ask for an additional$150 million in the 2001 budget to fight AIDS in Africa. Pfizer Inc. announced in April that it would donate the antifungal fluconazole, an effective treatment for cryptococcal meningitis—a lethal, opportunistic infection of AIDS—to poor South Africans. A few weeks later, World Bank president James Wolfensohn declared that the bank will place “no limit” on the amount it spends on AIDS. “I have said to our African clients, if you have programs, we'll fund them,” Wolfensohn declared. On 10 May, Clinton issued an executive order that relaxes U.S. laws regarding intellectual property rights connected to anti-HIV drugs in sub-Saharan Africa, effectively allowing countries to manufacture or import drugs without fear of violating patent law. Five pharmaceutical companies immediately promised to slash prices of their anti-HIV drugs sold in the region.

View this table:

On top of these commitments, several groups have donated money to help Africans take advantage of arguably the most meaningful advance yet made by AIDS researchers: the finding that relatively cheap, short courses of anti-HIV drugs can slow transmission of HIV from mother to child. With help from the Turner Foundation, the United Nations Children's Fund has started pilot studies of such interventions in eight countries in sub-Saharan Africa. France, aided by Glaxo Wellcome (which markets three anti-HIV drugs), last year started a similar program in Côte d'Ivoire called the Fund for International Therapeutic Solidarity. More support for drugs to thwart mother-to-infant transmission came in March from the Pediatric AIDS Foundation, which announced that it will award $800,000 in grants to seven programs in five African countries. As sorely as Africa needs this assistance, it will have only a modest impact on the continent's AIDS crisis in the immediate future. “Every war needs a war chest, but that provided by the international community is woefully empty,” Wolfensohn told the U.N. Security Council in January. Wolfensohn pointed out that Africa receives a mere$160 million now for HIV prevention; the bank estimates that Africa needs between $1 billion and$2.3 billion.

Public health experts stress that money is only part of the equation. Richard Feachem, director of the Institute of Global Health—a joint program of the University of California, San Francisco, and UC Berkeley—contends that African governments themselves have responded “halfheartedly” at best to their AIDS problems. Too many leaders, he says, have exhibited “ostrich behavior,” have been “sitting on their hands,” or have even actively made the situation worse. “If you liken this to a war, there are countries under attack losing huge numbers of individuals and doing nothing to defend themselves,” says Feachem, an epidemiologist who formerly was dean of the London School of Hygiene and Tropical Medicine and an official at the World Bank. “In the absence of an effective government response, it's very difficult for others to get going.”

Feachem, who has worked extensively in sub-Saharan Africa, has what he calls “a stock list of no-brainers” for governments: Educate the citizenry, provide condoms, clean the blood supply, work intensively with high-risk groups, provide early treatment for sexually transmitted diseases, and adequately train health providers. And he cautions researchers not to become “mesmerized by microdecisions,” particularly when it comes to treatment options. “People can argue in Durban for days and days about what drugs to use and how, but if they return to countries that are not mobilized for war, anything they do will be trivial,” says Feachem.

With similar impatience, Kevin DeCock, an AIDS epidemiologist with the U.S. Centers for Disease Control and Prevention in Atlanta who in 1988 started a large AIDS research program in Côte d'Ivoire, says that “the cost of the drugs is the least of the issues.” DeCock, who's now based in Kenya, believes prices will fall, and says many international companies working in Africa already can afford to subsidize treatment for their workers. But African countries can't easily identify HIV-infected people and provide them anti-HIV drugs, which require intensive treatment regimens. “We could end up spending a lot of money with very little benefit,” says DeCock. But he adds that “if the benefits of treatment are sustained in developed countries, it's increasingly going to dominate discussion.”

## To the front

No quick fixes will rescue Africa from the epidemic of HIV and AIDS that's already filling graveyards, crippling economies, and leaving legions of children to grow up parentless. But research does hold out the hope of, at the very least, optimizing the battle plans and, at the very best, forging the weapons themselves.

There are risks, however, in increasing research in sub-Saharan Africa, a region that for centuries has seemed to trade in one woe for another. For one, researchers themselves face danger; they operate on the front line, and an eruption of warfare can put lives in peril or destroy years of work overnight. Major AIDS research projects in the former Zaire, Rwanda, Sierra Leone, and Guinea-Bissau came to a sudden end because of civil wars. A coup took place last year in Côte d'Ivoire, northern Uganda remains embroiled in civil conflict, Zimbabwe has had deadly confrontations between poor blacks and wealthy white landowners, and Ethiopia and Eritrea are at war again. Although AIDS research projects continue in these locations, the investigators must keep a watchful eye on the political winds.

In peaceful areas there are other obstacles to overcome, such as the legacy of mistrust left by some past scientific studies. “To a large extent, colonial powers asked research questions that closely mimicked questions of interest in their country of origin,” says Malegapuru William Makgoba, head of South Africa's Medical Research Council. “In the old paradigm, people did research without explaining to people why it was good for them, and usually it was best not for participants but for researchers. Now it's incumbent on researchers to do research in a very different way and to make sure it's applied and relevant to [African] society.”

Translating research, ethics, the cost of drugs, politics, the spread of HIV, and the inexorably increasing toll of the AIDS epidemic all will receive a thorough airing at the 5-day Durban meeting that will take place 2 weeks from now. The 12 international conferences that have preceded this one focused largely on the concerns of wealthy countries. But HIV's impact on Africa surely will take center stage in this gathering.

Echoing the reaction of Milan's Clerici to the bleak morning tour in Uganda, UNAIDS director Piot thinks South Africa's intense epidemic of HIV will offer visiting delegates from developed countries their first chance to connect the mind-boggling statistics to individuals, communities, and a devastated society. (New figures suggest nearly a quarter of the country's adults are infected with HIV.) “I hope the conference will attract the brightest scientists in the West, so they'll be confronted with the real AIDS epidemic,” says Piot. “Whether you do sophisticated molecular biology or research with the latest antiretrovirals, I think it's really important that you get a sense of what the epidemic is all about and where it's going. I can't see how you can work on AIDS without having a burning curiosity to see what's going on where the epidemic is today.”

15. # Is AIDS in Africa a Distinct Disease?

1. Jon Cohen*
1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl.

Photos by Malcolm Linton

Opportunistic infections, transmission patterns, and viral subtypes differ, but there are more similarities than differences in HIV's behavior in Africa

Researchers and policy-makers around the world groaned when they learned of the fiery letter South African President Thabo Mbeki sent world leaders on 3 April, defending his decision to seek advice from “dissidents” who question whether HIV causes AIDS. Most dismissed Mbeki's infatuation with contrarian thought as dangerously misguided (see p. 2168). But his letter did highlight a fact that few AIDS researchers would challenge: “It is obvious that whatever lessons we have to and may draw from the West about the grave issue of HIV/AIDS, a simple superimposition of Western experience on African reality would be absurd and illogical,” wrote Mbeki.

Although AIDS researchers have solid data that suggest anti-HIV drugs work as well in Africa as anywhere else—removing a cornerstone of Mbeki's argument—many share the South African president's conviction that the African AIDS epidemic has unique features. As Mbeki noted, homosexual spread of HIV accounts for the majority of AIDS cases in the United States and Europe, while Africa has a primarily heterosexual epidemic. HIV also has spread more quickly in Africa, where different strains of the virus predominate. And AIDS-related diseases—and possibly disease progression itself—differ in Africa, too.

Yet 2 decades into the epidemic, AIDS researchers have only a shaky handle on the fundamental question: What accounts for these differences?

The answers have potentially important implications. “We'd like to know whether we could develop appropriate strategies for AIDS patients,” explains Gaston Djomand, head of the clinical section at Projet RETRO-CI in Abidjan, Côte d'Ivoire. “It makes a difference to know whether patients die 2 years after diagnosis of AIDS or 5 years. You can develop strategies to prolong healthy lives.” It also makes a difference to know whether the HIV strains circulating in Africa or the immune systems of people living there have special characteristics, two other topics that have attracted much attention—and debate.

## Location, location, location

The most obvious difference is clinical, and you don't need a medical degree to see it. Lacor Hospital in Gulu, Uganda, has a ward with a fenced-off courtyard, where many patients spend their days sitting with their families on bamboo mats, escaping the heat under blooming jacaranda and tulip trees. The courtyard has a graceful and peaceful feel, like a tuberculosis sanitarium from the past century. In fact, it is a TB sanitarium, but it is also largely an AIDS ward: Roughly 50% of the 100 patients, who stay for 2 months of treatment, are also infected with HIV.

Tuberculosis kills more HIV-infected people in Africa than any other AIDS-related disease. A study by the RETRO-CI research team, for example, found TB in 40% of HIV-infected people upon autopsy. In contrast, TB remains relatively rare in AIDS patients in the United States and Europe. Eleizer Masliah, a neurologist and pathologist at the University of California, San Diego (UCSD), recently found no TB at all in 390 autopsies of people who died from AIDS. “It's kind of surprising here, next to the [Mexican] border, where you'd expect to see more cases,” says Masliah. Similarly, Kaposi's sarcoma, a cancer that causes purple skin blotching, commonly afflicts HIV-uninfected and -infected people in Africa, while in industrialized countries it's largely restricted to HIV-infected, gay men.

Pneumocystis carinii pneumonia (PCP) shows the opposite pattern. Although malnourished African children have long suffered from PCP and many Africans have antibodies to the protozoa—indicating that it is common there—the RETRO-CI researchers only found P. carinii in 8% of the HIV-infected people autopsied. Similar results came from autopsy studies in Uganda and the former Zaire. Yet during the 1980s, PCP infected more than 80% of AIDS patients in developed countries at some time during the course of the disease, and the UCSD team found the protozoa in 30% to 40% of patients they autopsied. (Thanks to effective treatments for PCP and HIV, PCP rates in the late 1990s declined dramatically in developed countries; the UCSD study found the protozoa in only 7.6% of the 1998 autopsies.)

Although the contrasts may be starker than expected, researchers say the overall pattern is not surprising. “To some extent, what you see in HIV-positive people reflects what you see in the general population,” says James Whitworth of the U.K.'s Medical Research Council, which funds the AIDS program he co-directs with Anatoli Kamani at the Uganda Virus Research Institute in Entebbe. Tuberculosis, he notes, is much more common in Africa, and although he sees PCP autopsy data as “a big conundrum,” he suspects that many clinicians might miss it because it's relatively difficult to diagnose in Africa.

Whitworth is equally skeptical of claims that in Africa the disease typically progresses more rapidly from infection to AIDS to death. In the United States and Europe before anti-HIV drugs existed, gay men on average developed AIDS in 10 years and died in 12. In Africa, however, studies have come up with different results. One study, by the University of Nairobi's Omu Anzala and co-workers, found that Nairobi sex workers on average developed AIDS 4.4 years after infection, but the same researchers found that the disease progressed much more slowly in infected, pregnant Nairobi women. The difference, they surmised, may have been a result of high rates of sexually transmitted diseases (STDs) among the sex workers.

Whitworth, Dilys Morgan, Martin Okongo, and other members of his team have found rates of disease progression in the general population in Masaka, Uganda, similar to those seen in developed countries. Eight years into the study, about 40% of the people had developed AIDS, with death occurring 9 to 10 months later. Whitworth concludes that AIDS in the Masaka district looks like AIDS anywhere else. “There are lots of nasty diseases around here,” says Whitworth. “But with decent medical care, HIV-infected people in Africa can live as long as people in Europe and the West did before the era of antiretrovirals.”

## Ease of infection

Perhaps the sharpest disparity of all is the relative ease with which HIV seems to infect Africans. In studies conducted in the United States by researchers working with the U.S. National Institutes of Health to prepare for vaccine trials, high-risk populations such as gay men have had new infection rates of 2% per year. In some African populations, new infection rates are six times higher.

A combination of higher rates of untreated STDs, limited prevention campaigns, and lack of access to health care in general may account for much of the difference. But researchers are looking into two other possible factors: whether African immune systems are more vulnerable to HIV, and whether strains of the virus circulating in Africa may be more infectious than those circulating in the United States and Europe.

In an unusual study, the University of Milan's Mario Clerici, Lacor Hospital's Matthew Lukwiya (see profile on p. 2159), and their co-workers compared immune cells from two groups of people who were not infected with HIV: Ugandans and Italians living in Gulu versus Ugandans and Italians living in Milan. They found stark differences by location, but not by genetic background.

The researchers reported in the December 1998 issue of the journal AIDS that the immune systems of people living in Africa appear to be more highly “activated,” presumably because they must constantly fend off a much broader range of diseases. And a higher level of activation should mean more CD4 cells—HIV's main target—circulating in the blood. “You get more targets, you keep the virus happy,” says Clerici. Whitworth's lab has found some evidence to support this idea, but he remains skeptical. “There are loads of hypotheses, and which ones hold water, I don't know,” says Whitworth.

Theories about the impact of viral differences on ease of infection are focusing on a strain of HIV-1 known as subtype C—one of 11 subtypes, labeled A to K, of the most common group of HIV viruses. Subtype B predominates in the United States and Europe, rarely showing up in Africa, other than in South African homosexuals. East Africa has mostly subtypes A and D, while west Africa has predominately A and a combination A/G virus. (HIV-2, a distinctly different type of the virus, is also endemic to west Africa, but it has spread to other parts of the continent and the world. With Harvard University's Max Essex and Phyllis Kanki, Souleymane M'boup of Cheikh Anta Diop University in Dakar, Senegal, has helped establish that HIV-2 causes AIDS much more slowly than HIV-1 does.) In southern Africa, which had little HIV-1 until the 1990s, subtype C accounts for most of the infections. “The fact that C is circulating so rapidly is fascinating, and I tend to think it's different,” says Bette Korber, who oversees the HIV database at New Mexico's Los Alamos National Laboratory.

One difference, seen in studies of isolates from Malawi and Ethiopia, is in the cell surface receptors that HIV docks onto. HIV initially attaches to both a CD4 receptor and a coreceptor called CCR5 to establish an infection. Typically, as a person's immune system wears down, HIV begins favoring another coreceptor, CXCR4, over CCR5. But subtype C rarely makes the switch. This means that at any given time, a person infected with subtype C will likely have more copies of the virus that dock with CCR5—the variant most capable of establishing an initial infection.

Essex also believes that subtype C may transmit more efficiently through heterosexual sex. In the January Journal of Infectious Diseases, Essex and co-authors describe a chemical messenger secreted by immune cells that boosts HIV replication, particularly that of subtype C. Because venereal diseases increase secretion of this messenger, the researchers write that “it is tempting to speculate” that these conditions may explain the explosive spread of subtype C.

Other researchers see these speculations as just that. Saladin Osmanov, who tracks subtypes for the Joint United Nations Programme on HIV/AIDS, suspects that subtype C's current front-runner slot may simply be part of the ebb and flow of the epidemic. “Viruses tend to shift and disappear and come back,” says Osmanov. “I don't like simple explanations.” Virologist Martine Peeters of the Institute of Research for Development in Montpellier, France, agrees. “There are almost no subtype C data,” says Peeters, who has worked extensively in Gabon, Cameroon, and Senegal. “It's too early to conclude anything about any subtypes.”

HIV and AIDS do have different characteristics in Africa, but most researchers, ultimately, never lose sight of the main similarity: HIV, everywhere it goes, destroys immune systems, cuts lives short, and devastates communities. In Masaka, Uganda, says Whitworth, the risk of death increases by 11 times in people who test positive for HIV. Yet Whitworth says he also appreciates the Thabo Mbekis of the world: “It makes us think about assumptions and be quite sharp in our arguments.”

16. # Balancing the Collaboration Equation

1. Jon Cohen*
1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl.

Photos by Malcolm Linton

An inside look at three veteran marriages between African AIDS researchers and foreign colleagues illustrates the benefits—and stresses—of partnership

When researchers from the University of Manitoba in Winnipeg, Canada, began seeing an unusual number of cases of genital ulcers in the late 1970s, they turned to colleagues at the University of Nairobi for help. Friendships grew, and in 1980, the Canadians sent a senior infectious disease fellow to Nairobi for a year to study genital ulcer disease, which is relatively common in Kenya. That visit sowed the seeds for the longest running and one of the most productive AIDS research collaborations in Africa.

The project quickly expanded to include a study of sexually transmitted diseases (STDs) among prostitutes in a teeming Nairobi slum called Pumwani. When STD guru King Holmes of the University of Washington, Seattle, joined the project in 1984, a junior fellow working with him, Joan Kreiss, suggested they look into the prevalence of HIV among the prostitutes. Holmes and Manitoba's Frank Plummer, who now presides informally over the collaboration, tried to talk Kreiss out of the idea. They had seen no AIDS cases in Nairobi and thought the work would lead nowhere. Fortunately, Kreiss persisted. “We found that two-thirds of the women in Pumwani were infected with HIV, which was a complete shock to everyone,” recalls Plummer. “That turned things around completely.”

High-profile AIDS papers began to pour out of the collaboration. They described the extent of the epidemic in Kenya and linked high risk of HIV transmission to genital ulcer disease, contact with prostitutes, and lack of male circumcision. The collaborators discovered—much to the amazement of AIDS researchers around the world—that about 5% of the prostitutes did not become infected with HIV despite repeated exposure, suggesting that their immune systems might hold important clues for AIDS vaccine developers. “Our collaborations have really yielded results that have benefited the world at large,” says the University of Nairobi's Elizabeth Ngugi, who started the outreach work with the Pumwani prostitutes.

Given that string of achievements, you might think the Kenya project is a smooth-functioning model of cooperation. But everybody involved will tell you that it has not been easy. Kenyan researchers constantly find themselves balancing their expectations and needs with those of their collaborators, who not only provide the bulk of the funding but also have better access to organizers of conferences, editors at journals, and the international media. “There are equity issues that make it difficult, and it's a constant tension,” says Plummer. “The collaboration is like a marriage,” adds Job Bwayo, chair of the University of Nairobi's medical microbiology department and head of the AIDS research effort. “You have to give and take.”

Stresses and strains are not unique to the Nairobi partnership. In a dozen projects Science visited across sub-Saharan Africa, tensions are part and parcel of collaborative efforts. Most revolve around equity: access to financial resources and facilities, participation, transfer of technology, self-reliance, training opportunities, and credit. Foreign scientists also on occasion have had to confront their African colleagues about using lab resources to conduct personal business.

Yet, in spite of the difficulties, international collaborations are essential for conducting research in sub-Saharan Africa, where few countries can afford scientific research on AIDS—or anything else, for that matter. Indeed, they account for most of the research on the continent (see p. 2156). Three of the longest running projects—the Nairobi collaboration, a Swedish-Tanzanian project, and a collaboration in Côte d'Ivoire funded by the U.S. Centers for Disease Control and Prevention (CDC)—provide different models of how contrasting research worlds can work productively together.

## Staying power

The fact that the Kenya collaboration continues to flourish and produce groundbreaking results after 2 decades suggests that the partners in this marriage are doing something right. Chemistry is an important factor—as it is in most partnerships—and many of the participants credit Plummer with creating the right ambiance. “Frank is a quiet person who doesn't care about quarreling,” says Bwayo. “He's been central to this working.”

Plummer, a gentle but determined bear of a man with a salt-and-pepper beard, first came to Kenya in 1981 as the second fellow sent out from Manitoba. His 11-month stay led to a dozen papers co-authored with Kenyans, Canadians, and a young Belgian: Peter Piot, who now heads UNAIDS, the Joint United Nations Programme on HIV/AIDS. “I loved the place,” says Plummer. “I couldn't wait to get back.” He returned in 1984 and has made it home ever since.

The partnership continued to expand, bringing in universities from Seattle, Ghent, Toronto, Oxford, and Rotterdam. One key to its success is that the international partners—who together spend $3.6 million a year on research and another$1.2 million on intervention and training programs—have brought critical resources to the effort. A clinic to this day provides free care and treatment to Pumwani prostitutes. The University of Washington, with funding from the National Institutes of Health's Fogarty International Center, has paid for 48 Kenyans to come to Seattle for advanced training, and a few more have studied at Oxford, Winnipeg, and Antwerp. Kreiss started new studies across the country in Mombasa. The collaboration also built and equipped a lab on the Nairobi campus that has state-of-the-art machinery for doing HIV virology and immunology. “We wouldn't be doing any research without our collaboration,” says Bwayo.

The outside collaborators' willingness to listen to the complaints and concerns of their Kenyan partners helps explain the collaboration's survival, too. On 31 January 1998, they got an earful. On that Saturday, 22 researchers from Kenya, the United States, Canada, Belgium, and the Netherlands held a retreat to discuss the collaboration's strengths and weaknesses. A report of that meeting offers a rare peek into the frustrations faced by African scientists—and their wealthier collaborators—all over the continent.

Collaborators, the Kenyans said, should try harder to incorporate Kenyans when they draft research proposals, manuscripts, and conference presentations. University of Nairobi staff who currently do not work with the project should have more opportunities to participate. Collaborators should more freely share their late-model vehicles and computers. Rather than shipping samples out of the country because of a lack of equipment, more technology transfer should occur. The Kenyans also asked their collaborators to share information more freely about their budgets and encourage, rather than criticize, use of the lab for university and even private business.

This soul-searching led to a revised “Statement of Principles” and a list of “expected benefits and mutual obligations.” It didn't resolve all the problems, and collaborators challenged many of the criticisms, emphasizing that they had their own frustrations. Yet the collaboration, which has published nearly 300 papers, continues to expand. Recently, the group has started research projects in new communities, launched a project to develop Kenya's first AIDS vaccine, and drawn up plans to construct a new wing at the university.

## Long-term survivors

In neighboring Tanzania, a different kind of collaboration between Swedish and Tanzanian researchers, the TANSWED HIV Program, has been flourishing for 14 years. It has resulted in nearly 250 published papers and abstracts. Like most long-standing African AIDS research projects, this one has moved from describing the epidemic to testing interventions to slow HIV's spread, including preparing cohorts for vaccine studies.

“It's a true partnership,” says Fred Mhalu, an infectious-disease specialist at Muhimbili University in Dar Es Salaam. “Both sides are benefiting.” Gunnel Biberfeld of the Swedish Institute for Infectious Disease Control, who is well known for her AIDS vaccine studies in monkeys, agrees. “It's not one-way,” Biberfeld says. “We learn a lot.”

In contrast to the relatively freewheeling Nairobi collaboration, TANSWED is tightly structured. The Swedish International Development Cooperation Agency provides most of the funding—$8 million since 1986. It has strict rules designed to encourage technology transfer and discourage empire-building. It prohibits either side from hiring full-time investigators, for example, and doesn't fund new labs for the project. “We don't believe in establishing things separately, but to reinforce what's there,” says Biberfeld. “Then it becomes permanent.” The project has also provided support for many young Tanzanian researchers to participate in a “sandwich” program that combines brief training in Sweden with studies in Tanzania—a model that aims to reduce the risk of students staying abroad. These policies draw mixed reviews from the participants. Biberfeld says the prohibition on Swedish researchers working full-time in Tanzania slows down research and discourages Swedish investigators from participating. But Mhalu likes the policy. It encourages more Tanzanians to do the fieldwork, he says, and it helps eke out scarce resources. Full-time Swedish investigators, he notes, would eat up a big chunk of the budget. “I've seen several other projects that had to shut down because of how much they spend on [foreign scientists],” says Mhalu. On the other hand, neither side likes the ban on hiring full-time Tanzanians. “Our project has 30 scientists who have to teach, do clerical work, and don't have full-time research assistants,” says Mhalu, whose salary and lab space is paid for by the Tanzanian government. Biberfeld stresses that, at the end of the day, “this is a Tanzanian project.” Not only do the Tanzanians have nearly twice as many principal investigators as the Swedes, Tanzanians routinely present the data at international conferences and receive top billing in publications. ## A tale of one city Across the continent in Côte d'Ivoire, a French-speaking country on Africa's west coast, two large, often overlapping AIDS research projects coexist in the bustling city of Abidjan. The U.S. CDC funds the larger one, Projet RETRO-CI, while France's National Agency for Research on AIDS supports the other. Projet RETRO-CI, in collaboration with the Ivoirian Ministry of Health, set up shop in 1988, studying basic epidemiology of both HIV-1 and the less understood HIV-2, which is largely confined to western Africa. The CDC researchers and their Ivoirian collaborators were soon publishing their findings in high-profile journals, including Science. “It was lucky initially that we asked simple questions,” says CDC's Kevin DeCock, Projet RETRO-CI's first director. Like the Nairobi collaboration, RETRO-CI has brought critical resources to the project, including establishing a topflight lab (see profile on p. 2157), a clinic that provides free health care to sex workers, and a training program for African researchers. RETRO-CI, unlike TANSWED, has 135 full-time staff, but CDC limits the number of U.S. scientists (there are now only two), and African scientists are encouraged to publish papers and present at meetings. Yet the initial reception in Côte d'Ivoire was mixed. “When we started, people said, ‘What are you doing, research is not very important,’” recalls Ehounou Ekpini, the project's deputy director. Project RETRO-CI staff had a mixed reaction themselves in 1994 when French AIDS researchers, fleeing the Rwandan genocide, established another program in Abidjan. But after some initial wariness, the two groups began working together. In 1998, they pooled data on mother-to-child transmission of HIV during the postnatal period, and last year, they published back-to-back complementary papers in The Lancet demonstrating that cotrimoxazole, a relatively cheap sulfa drug, can decrease disease and prolong the lives of HIV-infected people. CDC's Stefan Wiktor, who stepped down as Projet RETRO-CI's director in August, has come to see the relationship as mutually beneficial. “It looks funny for people to do similar studies in the same city, but it's worked out well,” Wiktor says. “And among Ivoirians, they don't view it as a problem.” Issa Malick Coulibaly, who until last year headed the country's National AIDS/STD/TB program, says the intense research interest in Côte d'Ivoire has transferred much technology to his country. “We need to have a good capacity for research,” says Coulibaly. “We don't need to take blood of people living here and send it to France and the CDC.” As much as African countries want to build their research infrastructures, ultimately, the collaborations in Kenya, Tanzania, Côte d'Ivoire, and elsewhere won't be judged by how much they strengthen research, but by how much the research findings help prevent HIV and AIDS in these countries. And on that score, everyone is still struggling. “One of the frustrations I felt at the end of my stay here was lack of translation of research findings to public health programs,” says Wiktor. “Research isn't enough.” 17. # Major International AIDS Research Collaborations 1. Jon Cohen View this table: 18. PROFILE # Building a World-Class Lab From Scratch 1. Jon Cohen* 1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl. Photos by Malcolm Linton Abidjan, Côte D'IvoireShortly after Cameroon first started broadcasting television programs in 1984, John Nkengasong, then a student at the University of Yaoundé, remembers watching a show that featured veterinary scientists working in a laboratory. “I didn't even know what research was,” recalls Nkengasong, the son of a Cameroonian subsistence farmer. Now Nkengasong, a virologist at Projet RETRO-CI here, runs one of the best equipped HIV labs in sub-Saharan Africa and has published dozens of papers helping to clarify properties of specific strains of the virus. Projet RETRO-CI, a U.S.-Ivoirian collaboration funded by the U.S. Centers for Disease Control and Prevention (CDC), hired Nkengasong in 1996 to build a virology lab from scratch. As Nkengasong leads a tour of his lab, it's clear that scientists working in wealthy countries might well envy him. Dangerous viruses like HIV must be handled in sealed and aerated biosafety rooms. Nkengasong has two such P-3 rooms, one for HIV-1 and, to avoid contamination problems, another for the less common HIV-2. He has late-model polymerase chain reaction machines to amplify small amounts of DNA, a procedure routinely done in developed countries to determine how much virus people have in their blood—their “viral load.” A new DNA sequencing machine sits on one table and a sleek new Apple iMac computer on another. “There's nothing we can't do here,” says Nkengasong. After graduating from Yaoundé in 1987, Nkengasong won a scholarship to study in Belgium with Peter Piot at the Institute of Tropical Medicine in Antwerp. “He's the first African I put in charge of a lab in Belgium,” remembers Piot, who now heads UNAIDS and has trained several African AIDS researchers. “I really had a lot of confidence in him and am proud of his accomplishments.” Nkengasong says he badly wanted to return to Africa after completing his Ph.D. in virology at the University of Brussels. “There are many people in the States who are better than I am,” he says. “I can contribute to the development of Africa here.” But he could not find an attractive job in Cameroon. “The structures in Africa are not always there to receive the brains that are trained,” he says. When the chance to head the virology lab at RETRO-CI came up, Nkengasong jumped at it. A key part of Nkengasong's current work involves measuring viral loads, which can provide information essential for treatment and for critical research topics such as how much HIV a woman has in her breast milk or in cervicovaginal secretions. Nkengasong and his co-workers also have analyzed the subtypes of HIV-1 and HIV-2 circulating in Abidjan, which may lead to better screening tests for these viruses and, separately, help vaccine developers (see p. 2165). Nkengasong worries about the project's reliance on “soft money” from the CDC. “I'm always wondering about the day CDC won't do this,” he says. “Can the Ivoirians sustain this?” Ultimately, he believes, the lab could survive on grant money. “In 2 or 3 more years,” he predicts, “I'll be as competitive as any lab in the world.” 19. PROFILE # Restarting Rwanda's Research Engine 1. Jon Cohen* 1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl. Photos by Malcolm Linton Kigali, RwandaWhen Etienne Karita left here in 1993 to study at the University of Brussels in Belgium, the AIDS research community in Rwanda was booming. Karita worked with a French-Belgian collaboration conducting large studies of mother-to-child transmission of HIV; a project run by researchers from Johns Hopkins University in Baltimore, Maryland, was carrying out similar studies; scientists from the University of California, San Francisco (UCSF), were investigating the course of the disease in infected people and the factors that aid heterosexual transmission of the virus; and the World Health Organization had selected Rwanda as one of four developing countries best prepared to stage efficacy trials of AIDS vaccines. “There were very good collaborations with Rwandan doctors,” remembers Karita. “It was a spirit of teamwork that was beautiful.” But in April 1994, ethnic tensions between Rwanda's Hutus and Tutsis erupted into one of the greatest atrocities of the late 20th century, a genocidal rampage that killed an estimated 800,000 people during 100 days of terror. When Karita returned home in July, he learned that both his parents and eight of his 10 siblings had been killed. His wife, who had remained in Kigali with their daughter, was shot in the leg but survived by hiding above a ceiling panel. “People were massacred without any defense,” says Karita, his rage about the world's inaction accentuated by his quiet tone. Expatriate AIDS researchers fled the country shortly before the killings started, untold numbers of Rwandan AIDS researchers disappeared or died, and looters wantonly destroyed labs. One of Africa's most vibrant scientific communities suddenly vanished. Karita took his wife and daughter back with him to Brussels, where in November 1995 he completed a master's program in molecular biology with Peter Piot, now the head of UNAIDS. “He's a brilliant guy,” says Piot. “He could definitely have stayed in Belgium and had a quiet academic career.” But Karita opted to go home. “The idea I had in mind was, ‘What can I do to help Rwanda rebuild the country?’ I considered AIDS a priority.” Karita, who was a relatively junior researcher before the genocide, was named head of the country's National AIDS Control Program. With help from the Luxembourg government, Karita finished construction of a lab that was begun before the genocide, and in July 1997 he became head of the Belgian-funded National Reference Laboratory for Retroviral Infections. Today, this handsome brick lab does most of the HIV testing in Rwanda, and Karita, in collaboration with Dirk van Hove, a Belgian tropical disease specialist who also returned, has begun to restart the country's research engines. “Through it all, he's continued to try and improve his skills and stay abreast of current events in research,” says Susan Allen, an AIDS researcher at the University of Alabama, Birmingham, who headed the abandoned UCSF project in Kigali. Karita now collaborates with Allen, studying the immunity of people identified by her project who have survived 14 years with the virus, the “long-term nonprogressors.” The United Nations Children's Fund and the Pediatric AIDS Foundation also have awarded him grants to start testing and treating infected mothers to prevent them from passing HIV to their babies. But the research engine is still missing critical parts. Few Rwandan doctors have time to do research, says Karita. International donors fear investing in the country, too. Still, Allen is preparing a new grant proposal for the National Institutes of Health. And at each AIDS meeting Karita attends, he finds a few more potential collaborators. Says Karita: “As we're moving ahead, more and more people are starting to get interested.” 20. PROFILE # The Risks--and Rewards--of War Zone Research 1. Jon Cohen* 1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl. Photos by Malcolm Linton Kampala, UgandaAt 2:00 a.m. on Good Friday in 1989, about 30 young rebels from the Lord's Resistance Army breached the gate of Lacor Hospital in Gulu, where Matthew Lukwiya worked as a clinician, and demanded drugs. A gatekeeper roused Lukwiya from bed, and he told the rebels that the Ugandan army was keeping the drugs in town. They threatened to take some of the elderly Italian sisters from the Comboni Missionaries, a religious group that runs the hospital. “I said, ‘If you want to take the sisters, take me,’” recalls Lukwiya. Shoving people with their guns, the rebels did just that, grabbing five nurses, too. “They said, ‘Until you pay a ransom, we won't give them back,’” says Lukwiya. “They were very rough, and we thought they were going to kill us.” After keeping the hostages in the bush for a week, the rebels released this kind-faced young doctor and the nurses unharmed. But their harrowing experience explains why few Ugandan doctors, to this day, work in Gulu. “People don't want to risk [going there],” explains Lukwiya. Gulu, about 4 hours north of Kampala by car, is in an “upcountry” region torn by armed conflict. Residents are fearful of rebel soldiers, who have raped girls and seized boys to serve as fighters. Thousands of townspeople flood each day onto the guarded hospital grounds at dusk for a safe night's sleep. So upcountry clinical researchers like Lukwiya, a native of the Gulu district, play a critical role—and win high praise from their colleagues. “Matthew Lukwiya is the best doctor I have met in Africa,” says Giuliano Rizzardini of the University Hospital in Milan, an AIDS clinician who lived in Gulu himself for several years during the mid-1980s and still travels there regularly. Lukwiya works in Gulu primarily because he wants to provide care for his own people. He also takes the risk because he believes research can help thwart HIV in a region that still has shockingly high infection rates—in contrast to other parts of Uganda that have successfully curbed HIV's spread. In 1999, for example, 35% of the outpatients at Lacor Hospital tested positive for HIV. “It's very important to work in Gulu because when AIDS research started, researchers all centered on Kampala,” says Lukwiya, who in 1996 established a long-term research collaboration at Lacor studying immunologic responses to HIV with the University of Milan's Mario Clerici. (Lukwiya himself moved to the capital city last year but intends to return to Gulu soon.) “This cooperation with the Italians up north has done quite a lot for the people and the hospital,” says Lukwiya. “There's a transfer of technology and a transfer of scientific culture.” With the Italian researchers, Lukwiya has published cutting-edge papers that may help explain why Africans are particularly vulnerable to HIV, and he says he has been able to offer his patients better care. “Our hospital is the only one in the country that has systematic computerized records,” says Lukwiya. “We've been able to improve the training of our staff. And we've been able to link ourselves with good research centers in Italy.” Lacor has also attracted international attention, including a 1997 visit by U.S. Secretary of State Madeleine Albright, who pledged to establish a$2 million endowment for the hospital (a promise as yet unfulfilled).

Lukwiya hopes that if the developed world better understands the devastation that HIV and AIDS cause in places like Gulu, it will work harder to provide treatments. “People are dying,” he says. “They're dying here day and night.”

21. # The Mother of All HIV Challenges

1. Jon Cohen*
1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl.

Photos by Malcolm Linton

Several studies have shown that cheap, simple treatments can reduce mother-to-child transmission. Applying these advances is another matter

Rachel I. begins to cry, her gaunt frame shaking as she tries to contain her grief. Rachel, 31, is sitting in a social worker's office at a clinic in Abidjan, Côte d'Ivoire, recalling how her life has been torn apart by AIDS.

A few years ago, Rachel, who asked that her last name not be used, lost her first child, a 20-month-old boy. She soon became pregnant again and felt sick herself, so, at her husband's insistence, she finally went to see a doctor. The visit stunned her: She was infected with HIV. When she told her husband, who she believes infected her, he exploded with rage. Fearing for her safety, Rachel moved in with her mother.

Through her doctor, Rachel learned of an experimental treatment that might help protect her second baby. Researchers in Europe and the United States had discovered that giving the anti-HIV drug AZT to infected women during pregnancy sharply reduces the chance that the virus will be transmitted to her child. A French-sponsored trial called DITRAME in Côte d'Ivoire and Burkina Faso, headed by François Dabis and Phillippe Van de Perre, was testing whether a simpler—and cheaper—version of the therapy would work in Africa, where between 25% and 35% of babies born to HIV-infected women become infected themselves. Rachel enrolled in the trial and soon heard the first piece of good news she had received in a long time: Her second child, a girl, had been born uninfected.

But tragedy quickly struck again. Like the vast majority of women in Côte d'Ivoire, Rachel breast-fed her daughter. Six months later, she learned that she, too, had become infected, presumably through breast milk. “I haven't been well informed,” says Rachel. With help from France's Fund for International Solidarity, Rachel began to treat her baby with the anti-HIV drugs ddI, d4T, and 3TC. But it was too late, says Rachel. Her daughter died at 27 months of age.

Rachel's story epitomizes the dilemma throughout sub-Saharan Africa, where nearly 90% of the half-million babies infected by their mothers each year are born. AIDS researchers are finding cheaper and simpler ways to slow the spread of HIV from mother to child, and more pregnant women, even in the poorest countries, have access to anti-HIV drugs and formula—thanks to the largesse of donors, discounts from industry, new trade laws, and the tenacity of individual clinicians (see profile on p. 2163). But even these advances are barely making a dent in the grim statistics, because just as researchers offer ways to clear one enormous hurdle—drug availability—they run smack into other ones, ranging from social stigmas that discourage testing to disinterest on the part of cash-strapped health authorities to a deeply ingrained culture of breast-feeding—often supported by government policy for otherwise sound health reasons.

“With mother-to-child transmission, many scientific questions have been answered,” says DITRAME trial researcher Phillipe Msellati, sitting in an Abidjan café one evening this March. “But now we have to implement things, and the question is, ‘How do you do it in reality?’”

## Status quo

On Old Mulago Hill in Kampala, Uganda, prehistoric-looking marabou storks stroll on the lawn behind a mustard-colored shack and a canvas tent that helped change AIDS research history.

This shack and tent served as the main clinic and waiting room for a study that involved more than 600 pregnant, HIV-infected women and was run jointly by researchers who work across the street at Makerere University and their collaborators from Johns Hopkins University in Baltimore, Maryland. Half of the women received AZT throughout labor and delivery, and their infants received two doses of the drug each day for the first week of life. In the other group, women in labor swallowed one pill of nevirapine, an anti-HIV drug that the body processes more slowly, and the newborns received one dose of the same medication during the first 3 days after birth. Nearly every woman breast-fed.

In July 1999, an independent group of experts monitoring the trial abruptly ended it because of a startling difference between the babies in the two groups. When they were between 14 and 16 weeks old, 25.1% of the infants in the AZT group had become infected, while only 13.1% in the nevirapine arm were HIV positive. “It blew us all away,” says Laura Guay, a Johns Hopkins pediatrician who has lived in Kampala off and on since 1988 and helped run the study.

Guay has since been blown away by a less encouraging statistic: Not a single African country yet offers this $4 regimen of nevirapine to HIV-infected pregnant women, nearly 4 years after the drug came on the U.S. market. And only Uganda, Kenya, and South Africa have licensed it for sale. “It makes no sense to me,” says Guay, sitting in her office, located in a new building (built with Johns Hopkins funds) up the hill from the clinic. “In my talks with several government representatives from countries around here, they raise a lot of questions: ‘We don't know about long-term toxicities, we don't know about drug resistance, so we're not going to do anything.’ It's incredible.” By contrast, she says, when 076 came out, wealthy countries worldwide implemented it within weeks. Guay's shorthand mention of “076” refers to a study that provided the first convincing data indicating that anti-HIV drugs could reduce mother-to-child transmission of HIV. A drug regimen tested in the study—which was sponsored by the U.S. National Institutes of Health, as was the nevirapine study—cut the transmission rate by nearly 70%. But it had next to no relevance to poor countries, as the treatment cost about$800 and required an intravenous drip of AZT during labor. Further complicating matters, most women took the drug for 11 weeks before birth and none breast-fed. Poor African women, in contrast, often don't seek medical help until they go into labor, and most breast-feed.

In the 5 years between the headlines that celebrated the 076 results and those from the nevirapine study, several cheaper, simpler versions of 076 had proven themselves. Trials in Thailand, Uganda, South Africa, Tanzania, and Côte d'Ivoire, including the DITRAME study that Rachel I. joined, all showed that a short course of AZT or a combination of AZT and 3TC in the last month of pregnancy could lower the risk of transmission by up to 50%. Nevirapine promises similar or better results—with a much cheaper and simpler regimen.

## The breast-feeding dilemma

A big question mark hangs over all of these studies: the risk of long-term breast-feeding. The PETRA trial—the largest mother-to-child transmission study to date, with 1792 enrolled—has reported data from only 6 weeks after birth; many African mothers breast-feed for the first few years of life. When continued breast-feeding is factored in, “the data will not be all that rosy,” predicts Fred Mhalu, a leading AIDS researcher at Tanzania's Muhimbili University, one of the PETRA sites.

There's little question that breast-feeding greatly increases the risk of transmission. A 6-year study, conducted in Nairobi and published on 1 March in The Journal of the American Medical Association, compared infected mothers who used infant formula—provided by the study—to those who breast-fed (Science, 11 February, p. 942). Of the babies born uninfected, the study showed that formula-fed children had 16% fewer HIV infections at 2 years of age than the ones who exclusively breast-fed (for 17 months, on average). “We showed that in an urban setting with potable water, poor HIV-infected women with very limited education can use formula, and their children clearly have a much better outcome,” says Joan Kreiss of the University of Washington, Seattle, who conducted the study with the University of Nairobi's Ruth Nduati and co-workers.

The simple implication is that clinicians should discourage HIV-positive mothers from breast-feeding. But, as the University of Nairobi's Dorothy Mbori-Ngacha acknowledges, what sounds simple is, in reality, “a very challenging intervention” in an African setting. Many governments and international aid organizations have long promoted breast-feeding. Not only does formula cost more than many people earn, but many have no access to clean water, increasing the risk that formula itself can cause life-threatening diarrheal disease. The Nairobi study indicated that formula can be used safely in some situations, but “safe water is a big deal,” says Mbori-Ngacha. “Without it, formula becomes something you don't even want to discuss.”

Even if clean water is available, there are fundamental barriers to bottle-feeding. “If you don't breast-feed your baby, you're announcing that you're HIV positive,” says the University of Manitoba's Frank Plummer, who has lived in Nairobi for the past 16 years. “If you're on a bus and your baby starts crying, people will start yelling at you to breast-feed the baby.” Indeed, given nevirapine's potential for cutting transmission rates, “to us, the major risk factor now is the stigma of not breast-feeding,” says Hopkins's Guay.

## Social barriers

In the end, the social stigma of being HIV positive presents the most intractable problem health workers face in trying to prevent infected African women from transmitting the virus to their babies. Many women, fearing the angry reaction Rachel's husband displayed, don't even tell their partners their HIV status—even though many are infected because of their partner's infidelity. Such fears discourage many women from even being tested. “It's one of the areas we've neglected till now: sociobehavioral research,” says Thierry Roels, an official at Projet RETRO-CI. “We assumed if we found something, people would take it.”

The RETRO-CI study indicates that such an assumption is not well founded. In February 1998, researchers at RETRO-CI offered pregnant women in Abidjan free HIV tests and counseling. They offered those who tested positive free AZT at 36 weeks of gestation through delivery. More than 9600 pregnant women visited the clinic during this study, 72% of whom agreed to be tested. Thirteen percent tested positive. But of those, only 46% returned to learn the test results. And at 36 weeks of gestation, the researchers could find only 70% of the infected group. Most of these 117 women took AZT, but they represented a mere 12.6% of those who tested positive in the study. “One of the problems is [that] between screening and enrollment in the study, there's a long period,” explains Sibailly Toussaint, who heads the mother-child division of RETRO-CI.

With this in mind, the RETRO-CI team developed a rapid test that determines HIV status in hours rather than weeks. This enabled the researchers to test women and enroll them for treatment on the same day. Of the nearly 4000 pregnant women offered the test, 68% took it and 10.9% were positive. More than 75% of the women stuck around to learn their results—a big improvement. But half never showed up for treatment. In the end, only 37 of 294 infected women—again, 12.6%—took AZT.

Toussaint remains optimistic, though, that a simple regimen of one dose of nevirapine to mother and child might vastly improve these numbers. “Imagine using the rapid test and we give them the pill and tell them to take it when they go into labor,” says Toussaint. “We think we could reduce transmission significantly.”

## D-Day

Toussaint's optimism may soon be put to the test. On 20 March, Guay and Makerere University's Philippa Musoke, another principal researcher in the nevirapine study, excitedly watched a small truck pull into their parking lot on Old Mulago Hill. Inside was the first delivery of nevirapine sent to Africa as part of a project that will offer infected, pregnant women the drug free of charge, without even enrolling them in a study.

Similar nevirapine deliveries—part of an $800,000 effort funded by the Pediatric AIDS Foundation—soon will arrive in Rwanda, Kenya, South Africa, and Cameroon. Pharmaceutical companies have cut prices. Special projects launched by the U.S. and French governments will attempt to reach and treat pregnant women throughout Africa. Clinician Eric Mercier oversees an ambitious$4 million project funded by the United Nations Children's Fund, with AZT donated by Glaxo Wellcome, that also delivers drugs and formula to eight African countries. Mercier says the world must provide treatments as quickly as possible or pregnant women will take worthless medications. “There are plenty of fake drugs around the world, and women don't want to infect their children,” says Mercier. “They'll take anything.” In the same breath, Mercier urges patience, too. “We're in a world where people want to see immediate change and results. It's important to start now: People will use it, [and] others will see it and come into it. But it takes time.”

22. PROFILE

# Preventing Infections With Today's Tools

1. Jon Cohen*
1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl.

Photos by Malcolm Linton

With no vaccine in sight powerful enough to halt the spread of HIV, Africa is left with one strategy to combat the epidemic: prevention. But the record of most countries on this front is dismal. The University of Manitoba's Frank Plummer, who oversees a large research collaboration in Nairobi, Kenya, says his group has identified a half-dozen effective interventions—from bottle feeding to targeting heavily infected groups to male circumcision—to lower HIV transmission rates, but “the scale at which they're implemented is tiny.”

“We know how to prevent this infection by targeting behaviors and individuals who are at risk,” says Plummer, a specialist in sexually transmitted diseases (STDs) whose team's pioneering studies have focused on sex workers, long-distance truck drivers, adolescents, and babies of infected mothers. “It's been shown to work in Thailand and by our group here. … [The measures] need to be implemented as quickly as possible.”

The collaboration's work with Nairobi prostitutes indicates the powerful multiplier effect of interventions aimed at “core groups” that infect many people. A study of more than 1000 sex workers published in the April 1991 issue of the journal AIDS by Plummer, the University of Nairobi's Elizabeth Ngugi and Jeckoniah Ndinya-Achola, and co-workers estimates that their interventions—such as peer-group education, HIV testing and counseling (with distribution of condoms), and treatment of STDs—prevented between 6000 and 10,000 new HIV infections per year among customers and their contacts. The researchers calculated that it only costs between $8 and$12 to prevent each infection. “We need to work with vulnerable groups like sex workers in a very humane way so we can reach people in that population whom they're infecting,” says Ngugi, the main liaison with these women. “They are husbands. They are sons. Their sons have wives. High school kids tell me they go for ‘breakfast’ to sex workers.”

The Kenyan group also has produced evidence that male circumcision may lower the risk of HIV infection. In 1994, Stephen Moses, Allan Ronald, Plummer, and co-workers published a sweeping review of 30 epidemiologic studies showing a strong correlation between high rates of HIV infection in Africa and populations that don't practice male circumcision. “If I had my way, there would be circumcision offered to all adult men in AIDS-endemic areas, plus newborn circumcision,” says Plummer, who nevertheless recognizes the tricky medical and cultural issues.

Although the Kenyan group and several others have shown a strong correlation between having an STD and the risk of being infected with HIV, two large studies in Africa have come to different conclusions on whether widespread STD treatments would cut HIV transmission rates. One study published in 1995 followed 8500 people in Mwanza, Tanzania, half of whom took part in an intensified STD care and treatment program, while the others received the standard care in the region. At the end of 2 years, there were 40% fewer HIV infections among the treatment group. Then last year, researchers working in Rakai, Uganda, surprisingly reported that their trial, which was nearly twice as large, found no connection between STD treatment and HIV infection rates.

AIDS researchers are passionately debating whether differences in overall HIV prevalences between the two cities—4% in Mwanza and 16% in Rakai—or differences in STD treatment strategies could explain the contradictory results. But in any case, researchers emphasize that it makes sense to treat STDs. As a commentary in The Lancet—which published both reports—noted when the Rakai paper appeared, “anything short of an immediate response and a sustained commitment to STD prevention and control programmes is unthinkable.”

Although Plummer can point to successes in Kenya applying his group's research results, he says that for the most part, prevention efforts have few powerful advocates. “Prevention isn't at the top of the research agenda,” says Plummer. He then quotes Allan Ronald, the Canadian who started the Nairobi collaboration: “One hundred years from now, they'll say, ‘They certainly underestimated this and put far too little into it far too late.’”

23. PROFILE

# An Ambassador of Research

1. Jon Cohen*
1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl.

Photos by Malcolm Linton

Nairobi, KenyaElizabeth Ngugi sits on a well-worn mattress surrounded by prostitutes in a downtown brothel. “We feel great love for this woman,” says one of the dozen prostitutes who have jammed onto the three beds in the room to meet with Ngugi (pronounced en-goog-ee), a dignified woman who declines to give her age but who clearly has seen it all. “I know I would never be alive today without her.”

Indeed, without Ngugi, the collaborative AIDS research program based at the University of Nairobi may not have survived either. Ngugi, who joined the collaboration as a nurse in 1984 and went on to earn a Ph.D. in social work at the University of Washington, Seattle, plays a role that is crucial for many AIDS projects throughout Africa—but one that seldom receives credit in scientific circles. Like an ambassador of research, Ngugi connects ostracized communities that have little education and even less money to an international team of AIDS scientists that would like to follow them over time, learning the most intimate details about their sexual practices, their jobs and families, and, of course, their diseases. “With any infectious disease, you have to go where the prevalence is the highest in order to make a difference,” says Ngugi. “To make a difference with malaria, you have to go where the mosquito is breeding.”

The University of Manitoba's Frank Plummer, the de facto head of the collaboration, says the pathbreaking studies this group has done with these sex workers would have been impossible without Ngugi's help. “She's a public health phenomenon,” says Plummer. “She understands the country. She understands the values. And she's always reinforced that you can't just study things—you have to do something about it.”

In exchange for participating in studies, these women receive checkups by top-notch doctors, as well as free condoms, treatments for sexually transmitted diseases (STDs), and Ngugi's news from the front about the latest weapons to combat HIV. Today, Ngugi speaks with the women about the female condom, a vaginal pouch that protects against HIV transmission but has a reputation, deserved or not, of being bulky and cumbersome. The women, holding six conversations at once, quiet down when Ngugi takes the hand of a woman wearing a colorful scarf on her head. “Let her tell her story,” says Ngugi. “It's beautiful.”

Prostitutes often receive offers of extra money if they do not require a man to wear a condom (a sure sign, these women say, that the man is infected). The woman wearing the scarf explains how she used a female condom that Ngugi gave her to turn the tables on a customer. “Men never know you have anything in there,” the woman explains. “A few days ago, a man came and wanted to have sex, and he had never used a condom at all. I said, ‘Wait, are you willing to pay more money?’” So the man paid extra, not realizing that she had on a condom. Ngugi and the other women have a good laugh. “Let's not get rid of the male condom, but let's empower women and give them something to negotiate safe sex [with],” Ngugi declares.

Ngugi says the work she and her colleagues do “has paid off in bigger ways than we can measure.” But in one area—reducing the disease burden on the prostitutes she works with—there's at least an indirect measure of her success: “People say if we want to study STDs,” Ngugi notes, “we have to go where you're not.”

24. PROFILE

# McIntyre and Gray Have a Will and a Way

1. Jon Cohen*
1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl.

Photos by Malcolm Linton

Soweto, South AfricaIn an expansive waiting room for pregnant women at the Chris Hani Baragwanath Hospital here, one sign shows a drawing of a pistol inside a red circle with a slash through it. An embossed placard next to a check-in desk reads, “PATIENTS WILL ROUTINELY BE CHECKED FOR AIDS.” A flyer taped to another wall by the Perinatal HIV Research Unit states in bold red type, “NO HIV POSITIVE WOMAN SHOULD GIVE BIRTH WITHOUT ANTIRETROVIRAL TREATMENT.”

These three signs, each radical in its own way, capture the complicated world where pediatrician Glenda Gray and obstetrician James McIntyre aggressively try to help HIV-infected pregnant women stop the virus from infecting their babies. The international symbol for “no guns” hints at the level of mayhem just outside the razor-wire-topped walls that protect Chris Hani from the infamous shantytown of Soweto. The notice that patients will routinely be checked for AIDS—something that has never been done to whites—is a relic, a reminder of how badly the apartheid regime violated blacks, who make up nearly all of the patients here. And the sign about antiretroviral treatment is not simply a rallying cry but, rather, an amazing offer in a country whose government has declined to provide anti-HIV drugs to infected pregnant women (see p. 2168).

At Chris Hani, which suffers the dubious distinction of having more beds (3300) than any other hospital in the world, HIV infects more than 20% of the 17,000 pregnant women who give birth here each year. Gray and McIntyre constantly scout for appropriate clinical trials, new programs launched by nonprofit organizations such as the United Nations Children's Fund and the Pediatric AIDS Foundation, and anyone with deep pockets, which explains how they've managed to stockpile enough anti-HIV drugs to offer free treatment to infected pregnant women. According to McIntyre, 90% of the women they see accept the offer of free HIV testing and counseling, and virtually everyone who tests positive requests treatment. And when informed about the risk of transmitting the virus via breast milk, more than 90% of these women purchase formula that the clinic sells at a discount, says McIntyre.

Gray emphasizes that researchers can make a huge difference simply by applying pressure on drug companies and donors. She notes that UNAIDS even supplied them with AZT and 3TC long after the completion of a clinical trial so they could continue to treat women who had been helped by the drugs. And Gray is livid about her government's lack of a treatment policy. If a white government behaved that way, says Gray, “there would be a revolution.” Costa Gazi, health secretary of the Pan African Congress, who has harshly criticized the ruling African National Congress's AIDS policies, says the country owes a “massive debt” to Gray and McIntyre.

Walking through the waiting room, McIntyre pauses to look at the apartheid-era warning that everyone will be tested for AIDS. “We really should take that down,” he says. And he worries that they may run out of anti-HIV drugs one day and have to take down the sign they have taped to the wall, too.

25. # Searching for the Epidemic's Origins

1. Jon Cohen*
1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl.

Photos by Malcolm Linton

Many theories purport to explain how HIV got a foothold in humans, but few researchers have gone into the field to look for answers

Eastern GabonDriving through the jungle on a brick-red mud road, Paul Telfer suddenly stops his Land Cruiser. “I saw a cephus,” he says. Telfer backs up 100 meters and, sure enough, a dead monkey lies atop a metal drum, a dead brush-tail porcupine by its side.

In this remote part of western equatorial Africa, bushmeat makes up a large component of the diet. Hunters walk about with shotguns and machetes and display their bounty to the few cars that pass by in hope of making a sale. Telfer does not want to buy the monkey, though. He is looking for viruses. A primatologist from California, Telfer does fieldwork for the only team in the world that hunts for new simian immunodeficiency viruses (SIVs), HIV's cousins. He and his collaborators have found SIVs in chimpanzees, sooty mangabeys, mandrills, red cap monkeys, and two dozen other species. “Our goal is to understand the evolution of these viruses and attempt to understand what caused the AIDS epidemic and [what] the risks of future epidemics [are],” says Telfer.

Theories about the origin of AIDS all suffer from a dearth of data, and watching Telfer work explains why: It is labor-intensive and often tedious work. Today, Telfer is making a 4-hour round trip from his home base at the International Center for Medical Research of Franceville (CIMRF) to the village of Okondja solely to draw blood from a pet Cercopithecus cephus he found on an earlier expedition. If he can sample some dead animals along the way, so much the better.

Telfer, a Ph.D. student who works for virologist Preston Marx of Tulane Regional Primate Research Center in Covington, Louisiana, has spent much of the past 9 years cruising around Sierra Leone and Gabon searching for SIVs; few scientists have his knowledge or skills. Working with his Gabonese translator and assistant, Issa Bedjabaga, Telfer immediately displays a critical talent: the ability to convince someone selling a dead monkey to let him take samples for free. (He refuses to pay because he does not want to encourage the bushmeat trade.)

A woman comes out from her shack and gives Telfer and Bedjabaga her OK. After donning plastic gloves, Telfer plucks a few hairs, from which he later will extract DNA to confirm the species. He then weighs the animal and takes many measurements, with Bedjabaga recording the data in a notebook. When Telfer cuts off a bloody piece of the monkey's tongue, many men who have come out from nearby shacks begin talking excitedly. Telfer, who now is drenched in sweat, would like to remove the animal's spleen—a rich source of the white blood cells that SIV infects—but Bedjabaga, sensing that the men think Telfer is damaging their goods, hurriedly says goodbye to the villagers.

Telfer has no idea whether this species carries SIV or whether he can isolate it from a bloody tongue sample, but he is thrilled. “That's priceless data, priceless data,” he says, driving away. “You can never pass up an opportunity, because you never know what you're going to find. We may end up finding an important link.”

For more than a decade, researchers have recognized the close genetic links between SIV from sooty mangabeys (SIVsm) and HIV-2, a type of the virus that is mostly confined to west African countries. Pioneering work by Marx, Telfer, and co-workers at the Aaron Diamond AIDS Research Center in New York City (Marx's other affiliation) further matched SIVsm found in wild animals to HIV-2 infections from the same area.

To date, there are scant data that closely link HIV-1, the predominant human AIDS virus, to SIVs found in primates. The closest connection comes from viruses, called SIVcpz, isolated from a mere six chimpanzees. France's François Simon, who works at CIMRF and collaborates with Telfer and Marx, co-authored a paper published in the January Journal of Virology that strongly ties HIV-1 found in humans in Cameroon to SIVcpz isolated from two chimpanzees. But these isolates are most closely related to the least common of the three known HIV-1 groups, a group known as N; they do not, then, explain the origin of the AIDS epidemic, which consists mostly of M (or main) group viruses.

Studies from the lab of another CIMRF collaborator, Beatrice Hahn of the University of Alabama, Birmingham, last year placed three of the other four chimp isolates somewhere among groups M, N, and O. Most interestingly, the researchers discovered that these three viruses came from one of four chimpanzee subspecies, Pan troglodytes troglodytes. The remaining isolate—which is different from the others—came from the schweinfurthii subspecies. Because chimps do not typically cross rivers, the subspecies have unique ranges, which led the researchers to conclude that HIV-1 initially jumped from chimps to humans in a restricted area of west equatorial Africa that includes Gabon. Most AIDS researchers suspect that the virus crossed the species barrier through the butchering of primates.

All of these details became a topic of hot debate last fall upon the publication of The River, a thick book by former BBC correspondent Edward Hooper, who traces the origin of AIDS back to trials in the late 1950s of an oral polio vaccine in Congo, Rwanda, and Burundi. Hooper confirms many early cases of AIDS in that region and posits that the polio researchers, who had a chimpanzee colony in Congo of the schweinfurthii subspecies and Pan paniscus (a separate species known as bonobo) common in that area, used cells from the animals to make the vaccine. According to the theory, some of these animals had SIVcpz infections with viruses more closely related to the M family than anything yet found. Their infected cells, Hooper argues, contaminated the vaccine. “I find it remarkable that there's a dearth of M present [in humans] in any troglodytes countries earlier than the 1980s,” says Hooper.

There is no evidence that those who made the oral polio vaccine used chimp cells, however, and they have forcefully denied the assertion. The CIMRF collaborators also, to a person, reject the polio vaccine theory. “The two events have no connection,” says Simon. “It is as stupid as to say, ‘Since Africans drank Coca-Cola for 50 years, they have AIDS.’”

Hooper has had some strong scientific supporters, most notably the late Oxford evolutionary biologist William Hamilton. In January, Hamilton took a team to eastern Congo, hunting for SIVcpz samples from schweinfurthii and Pan paniscus. Hamilton, who collected samples of chimp feces to test for the virus, died a month later from malaria he contracted during the trip. The samples are now being analyzed. Hooper also has identified many labs that have old African samples of blood or tissues, and he has lobbied them to test the samples for HIV. His earlier efforts paid off when Aaron Diamond's David Ho tested a 1959 blood sample from a Congolese man and found HIV-1 sequences, now the oldest confirmed HIV-1 sequences on record.

Ratcheting up the debate, Bette Korber and colleagues at Los Alamos National Laboratory in New Mexico reported in the 9 June issue of Science (p. 1789) that their computer analysis of various HIV-1 isolates dates the origin of the M group to between 1915 and 1941—long before the polio vaccine tests took place. “Our results don't disprove that hypothesis but make it unlikely,” says Korber. In September, the U.K.'s Royal Society plans to host a meeting where Korber and other leading AIDS researchers—and Hooper—can debate their ideas about HIV's origins.

Telfer and colleagues hope soon to begin a systematic collection of chimp feces in Gabon. Telfer will also continue to sample as many primates as he can find: On this day, he not only bleeds the pet monkey, he also takes tissue from yet another dead cephus. “We need to know how HIV-1 and HIV-2 occurred,” says Telfer. “We need to understand not just the ultimate causes but [also] the proximal causes to start thinking about an HIV-3 epidemic.”

26. # Africa Boosts AIDS Vaccine R&D

1. Jon Cohen*
1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl.

Photos by Malcolm Linton

Efforts have recently heated up in several countries to tailor-make preparations that many believe offer the best hope yet for stopping HIV cold

Hlabisa, Kwazulu-Natal ProvinceThis lush region of gentle hills dotted with round mud huts that sits on the edge of the Hluhluwe Game Reserve doesn't look like an epicenter of South Africa's AIDS epidemic. But this poor rural area has one of the worst outbreaks of HIV in the world. The figures are staggering: Almost 40% of the pregnant women—a so-called “sentinel population” widely used to track the epidemic—in Hlabisa are infected with HIV. A decade ago, almost none tested positive.

In the sad calculus of AIDS, this makes Hlabisa an ideal place to test HIV vaccines—the best hope Africa has of truly containing this disease. And that was the reason Seth Berkley, head of the International AIDS Vaccine Initiative (IAVI), paid a visit here in late March.

Concerned that industry has shown little interest in developing an AIDS vaccine for poor countries, the nonprofit IAVI has launched innovative “product development teams” in South Africa, Kenya, and Uganda. Here, IAVI has teamed leading South African AIDS researchers with a North Carolina biotech company; together they hope to tailor an AIDS vaccine by next year for trials in Hlabisa (pronounced shla-bee-sa) and elsewhere in the country. As part of the project, to which IAVI has committed $4.5 million over 3 years, researchers work closely with participating communities. Today, eight “community educators,” whose job it is to lay the groundwork needed to conduct an ethical trial, pack into a trailer that serves as the headquarters of the Hlabisa Research Clinic to meet this boyish and gangly physician from New York. Ethics—something often ignored in the checkered history of medical research in Africa—demand that people who volunteer for the trial do so freely, fully understand whatever risks exist, and receive encouragement and advice about how to avoid becoming infected. These community educators, college-age men and women from the area, go from hut to hut to teach people about vaccines and AIDS, distribute condoms, and answer questions. Berkley begins by asking the educators, “What's the community saying?” “They feel positive here,” says a university student, who wears a stylish beret. “They think they're going to help stop the epidemic.” “Do they ask any questions that stump you?” Berkley queries. The educators squirm a bit. Berkley asks the question again. “The one is, ‘When are we going to have a vaccine that works?’” a woman answers. “That's a good question,” Berkley replies. The educators laugh. Berkley, who well understands the scientific and ethical challenges ahead, does not offer an answer. ## Off safari Human trials of AIDS vaccines began in Africa. The first test was conducted in Kinshasa in 1986 by Daniel Zagury of France's Pierre and Marie Curie University and his colleagues in Zaire. Ultimately, the trial went nowhere and was widely criticized for being run without the international community's knowledge, involving children, and using unapproved materials in the vaccine. No evidence ever surfaced that Zagury's trials injured anyone or violated Zairian or French regulations, and they did establish that a person could safely develop immune responses against pieces of HIV, which all vaccines contain. But the trials—which triggered media investigations and an official probe by the U.S. National Institutes of Health (NIH), which exonerated Zagury—also put AIDS vaccine researchers on notice that the ethics of the trials they stage, especially in poor countries, would receive intense scrutiny. No other vaccines were tested in Africa until last year, when Uganda started a small trial to assess the safety of an Aventis Pasteur product, which had already been widely tested in developed countries (see table). Not only have most companies shrugged at the African market, but government agencies such as NIH have opted to conduct initial tests in developed countries, for ethical and practical reasons. “When in doubt with a brand-new candidate vaccine, you'll cause the least stress if you try the vaccine out in Bethesda, Maryland,” says Yale University's Robert Levine, a bioethicist who has helped craft two sets of guidelines covering AIDS vaccine tests. View this table: The first of these guidelines, a 1993 document prepared by the Geneva-based Council for International Organizations of Medical Sciences (CIOMS), said initial studies of a vaccine “should be conducted only in developed communities of the country of the sponsor.” Levine says the guidelines aimed both to prevent “safari research”—in which researchers bag data in Africa to support licensing of products back home—and to protect more vulnerable populations from early safety studies that could take place anywhere. This year, Levine helped craft new guidelines by UNAIDS, the Joint United Nations Programme on HIV/AIDS, that relax this edict, concluding that developing countries can dive in first if they so choose. “Attacks on [the old guideline] came from low-income countries,” says Levine. “They asked, ‘Why is it that you treat us so paternalistically?’” The new guidelines reinforce the CIOMS principle of “equitable distribution,” which says communities that participate in trials should have access to the vaccine if it proves effective. IAVI follows that principle by insisting that the vaccines it bankrolls sell for a “very reasonable price,” which the nonprofit hopes to regulate. First, IAVI requires companies it works with to reveal production costs and then allows them to make a negotiated profit (10%, in one case). If IAVI deems the company's production costs too high, it can hire a third-party manufacturer. “IAVI's access to patents and in some cases the manufacturing know-how will certainly be helpful,” says Peggy Johnston, who heads the AIDS vaccine program for NIH's National Institute of Allergy and Infectious Diseases (NIAID)—and who previously worked as IAVI's scientific director. But Johnston cautions that even these stipulations do not ensure community access, as it may ultimately involve a tangle of patent rights, with other manufacturers seeking licensing fees. Success with anti-HIV drugs forced the drafters of the UNAIDS guidelines to confront another contentious issue: No vaccine works 100% of the time, so should sponsors provide the “best proven therapy” in the world to participants who become infected during the trial, or simply “the highest level of care attainable in the host country”? After fierce debate—including fire fights in The New England Journal of Medicine and the U.S. Congress—UNAIDS decided that each country should answer this question for itself (Science, 3 July 1998, p. 22). ## Natural immunity Finding the recipe for a vaccine that works remains, of course, the biggest obstacle facing researchers. Nothing tested in humans to date shows great promise, but even a mediocre vaccine could help. A 1992 NIAID statistical analysis made this point dramatically, showing that over the course of a decade, a 60% effective vaccine introduced today would prevent nearly twice as many infections as one with 90% efficacy introduced 5 years from now. African studies, particularly in the IAVI-sponsored Kenyan project that Omu Anzala works with (see profile on p. 2166), are turning up critical leads that might help make vaccines work better. Agnes Monifa is providing one such lead. Monifa lives outside Nairobi in Pumwani, a shantytown of wattle and mud homes with corrugated tin roofs that also has a clinic, run by the collaboration, for sex workers. Monifa, who is 48 years old, says she started in the sex business in 1973. She was one of the first women to use the clinic's services, which freely treats the women's sexually transmitted diseases (STDs) and other health problems. Since 1985, doctors at the clinic have been filling test tubes with samples of Monifa's blood—some of the most interesting blood that AIDS vaccine researchers have ever studied. Monifa is one of a group of Kenya prostitutes who, despite repeatedly being exposed to HIV, show no evidence of an infection. Inside the dark space of her one-room shack, Monifa, a mother of five, describes the details of her life through an interpreter. On a good day, she'll have 10 customers who will pay 50 shillings—65 cents—each. Monifa insists that men wear condoms, she says, but the evidence suggests otherwise: The clinic has provided her with so many treatments for STDs that, she says, “if I had to buy my medicine, I'd be dead by now.” Several close friends, she says, have died from AIDS. To explain her own fate, she turns to religion. “God loves me,” she says. Surprise fills Monifa's weathered face when she learns that her blood might help researchers find a vaccine. “I didn't know that,” she says. “I'm ready to help people.” In the early days of AIDS vaccine research, most studies focused on antibodies, the Y-shaped proteins that glue themselves to invaders and prevent them from establishing an infection. Monifa has no HIV antibodies in her blood. But the Kenyan team has shown that many “resistant” women who visit the Pumwani clinic have white blood cells trained to kill HIV after it has entered a cell. In part because of this work, several vaccine developers have put an increasing emphasis on stimulating production of these killer cells. Yet this research also has uncovered less encouraging results. Recently, 10 of the 80 or so “resistant” women in this study became infected—after taking a break from sex work (Science, 11 February, p. 942). This odd finding challenges the robustness of “natural immunity,” the phenomenon of immunologic memory that protects people from being infected twice by the same bug. Peter Kiama, a Pumwani clinician, contends that memory sometimes is fickle, requiring repeated stimulation from HIV to maintain immunity. Fears have surfaced, too, that immunity to one strain of HIV might not protect against another. Logically, the simplest vaccine would train an immune system to defeat the strain of virus circulating in a given community. So IAVI and others insist that designers of vaccines, all of which contain modified forms of HIV, rely on subtypes circulating in the region where they intend to test the preparation. The small Uganda trial launched last year, however, did not follow this dictum, testing a vaccine based on subtype B—the most common subtype in the United States and Europe, but rare in Africa. David Serwadda of Kampala's Makerere University says this decision didn't sit well with some Ugandan researchers. They passed these concerns along to their collaborators. “We said, ‘You've got a better negotiating position if you put in a [subtype] that matters to us,’” says Serwadda. Although the Ugandans agreed to participate in that small trial, Aventis Pasteur now is customizing an AIDS vaccine from an HIV subtype circulating in Uganda. NIAID's Johnston, whose institution sponsored the Uganda trial, stresses that scientists still don't know the significance of the different subtypes. “If someone does know,” she says, “they have a crystal ball clearer than mine.” The University of Munich's Michael Hoelscher is trying to make the crystal ball a little clearer. He has organized a large, multinational team to study 600 bar workers in Tanzania, where subtypes A, C, and D “cocirculate.” Hoelscher says unraveling the success—or failure—of immune responses at stopping a second HIV infection may clarify their relationship to subtypes. “Nobody knows anything about dual infections,” says Hoelscher. “It's a totally new field.” ## Vaccine dreams Halfway into the meeting of the community educators in Hlabisa, two nurses who work at the local hospital walk into the trailer, taking seats next to Berkley. “If we had a vaccine that worked and was safe, would people take it?” Berkley asks the nurses, Catholic sisters who wear military-style shirts with epaulets. “People would flock to it,” says one sister. “We're all praying for it,” adds the second sister. “Our children are dying. Our friends are dying. People 18 years old are lying in hospital beds that in the old days we had filled with grannies who had tuberculosis and cardiac failure. The vaccine would be wonderful. [The situation] is very difficult for young people. This is the time when they should be falling in love, and they say, ‘Now that we've come to that stage, you want us to do what you did not do.’” Berkley thanks everyone for coming to meet with him and for helping to organize an AIDS vaccine trial in this remote corner of Earth. “This is not just a problem for KwaZulu-Natal,” says Berkley. “This is not just a problem in Hlabisa or South Africa. It's a problem for the world.” 27. PROFILE # A Native Son Comes Home to Help 1. Jon Cohen* 1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl. Photos by Malcolm Linton Nairobi, KenyaIn December 1996, when Omu Anzala returned to Kenya after completing his Ph.D. work in Canada on HIV's interactions with the immune system, his older brother was part of the family contingent that welcomed him home. “Just looking at him at the airport told me that all was not well,” remembers Anzala. “They never told me.” Four months later, his brother, then 37, died from AIDS. Anzala, now a University of Nairobi researcher who runs the Kenyan end of one of the most advanced AIDS vaccine projects in Africa, takes a long look at the tabletop in his sparsely decorated office. “With all this knowledge I'd learned, I couldn't do anything,” says Anzala. “I tried to get all the antiretrovirals, but it was late. It told me why I am even studying this. So AIDS is close to me. Very, very close. That really makes me push for vaccines, because that's where we'll save the majority of Kenyans and the majority of people.” Since 1987, Anzala has worked with a collaborative AIDS research project started by researchers from his university and the University of Manitoba. “If there's anyone who has benefited from this collaboration, that would be me,” he says. The collaboration helped him earn his Ph.D. in Canada, which in turn led to postdoctoral work at Oxford University with one of the world's leading HIV immunologists, Andrew McMichael. Now the International AIDS Vaccine Initiative has funded Anzala, McMichael, and two European companies to fashion an AIDS vaccine based on a strain of HIV circulating in Kenya. Kelly MacDonald, a veteran of the collaboration who studied with Anzala in Canada and now directs the University of Toronto's HIV research program, says Anzala represents “the new breed of African scientists” who are trained abroad but working at home with real opportunities. “There's a shortage of people like Anzala in Kenya and elsewhere in Africa,” says MacDonald. Anzala emphasizes that he has had to chart his own course. He says, for example, that he was encouraged to study epidemiology rather than virology for his Ph.D. “The Canadians told me that's what Africa needs,” says Anzala. “No. We've been counting numbers for a long time. You want me to keep counting numbers? For how long?” Anzala replied. He also had to convince his Oxford collaborators that the Nairobi lab, if properly equipped, could do sophisticated analyses of blood samples for the vaccine trial. “They wanted us to ship things,” says Anzala. “The whole issue has been, ‘Oh you don't have support, you can't order reagents.’” He eventually prevailed. “We can do much more than be a collection point for other labs,” asserts Anzala. If all goes well, trials of the Kenyan vaccine—which is designed to boost the body's so-called killer T cell defenses—will begin in Nairobi by December. Anzala stresses how badly his country needs this weapon. “The answer to HIV and AIDS infection in Kenya is not going to be condoms and education,” he says. “We've been telling people that for 15 years. What is happening? We're still seeing more infections. What we really need is the same thing that prevents measles and has almost eradicated polio.” 28. # South Africa's New Enemy 1. Jon Cohen* 1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl. Photos by Malcolm Linton Many South Africans long dreamed of the day when the oppressive apartheid system would end. That day has come, but now the country faces a new disaster: one of the world's worst HIV epidemics—and most confusing government responses Since the beginning of this year, an estimated 320,000 South Africans have become infected with HIV, accounting for fully half the new HIV infections in sub-Saharan Africa. During that time, South Africa's president, Thabo Mbeki, has been publicly questioning whether HIV is the cause of the AIDS epidemic that is ravaging his nation. He has established a four-person panel to come up with tests to prove whether the virus causes the disease, although virtually every AIDS researcher believes the cause was proven beyond any reasonable doubt in the 1980s. The panel plans to report its findings in time for the international AIDS meeting to be held in Durban next month. Mbeki himself is scheduled to deliver the keynote address, which should guarantee that his skepticism will continue to get headlines around the world. Mbeki describes his interest in discredited ideas about HIV as a search for a better way to deal with the epidemic in Africa. Outsiders may see it as a flirtation, but many AIDS researchers here see it as irresponsible. While HIV infection rates soar in South Africa, the government has stumbled from controversy to controversy, undermining confidence in its public health strategy. Although the Mbeki administration has supported AIDS prevention efforts and research, it has ignored its own experts' advice to provide AZT to HIV-infected pregnant women. It has become mired in scandals, like a former health minister backing a supposed wonder drug for AIDS that turned out to have no scientific worth. And last year, it even failed to spend nearly half of its AIDS budget. “It's not a good time to be a researcher in South Africa if your pursuit is science and truth,” says epidemiologist Mark Lurie of the Africa Centre for Population Studies and Reproductive Health, a Wellcome Trust-funded effort in Mtubatuba. ## Fertile ground Salim Abdool Karim, head of the HIV Prevention and Vaccine Research unit in Durban that's sponsored by the Medical Research Council (MRC), says the government's confused response to the epidemic, combined with a large migrant workforce, high rates of sexually transmitted diseases, a thriving sex worker industry, and widespread poverty, provided a perfect breeding ground for HIV. “You couldn't ask for anything else,” says Karim, who chairs the scientific program for the upcoming Durban meeting. “If I were going to design an infection that was going to ravage this country, I could not do better than HIV and AIDS,” says Karim, a large man who goes by the nickname “Slim.” South Africa has had two HIV epidemics. The first took place in the 1980s and remained largely restricted to homosexual men; the second, much larger one took off in the 1990s and spread rapidly among heterosexuals. Studies from prenatal clinics across the country, which form the basis for national estimates, show prevalence skyrocketing from 0.76% in 1990 to 22.4% by 1999 (see left-hand graph). South Africa's Ministry of Health estimates that HIV now infects 4.2 million children and adults in the country, about 10% of the population—and more than 10% of the total infections in the world. A strain of the virus called subtype C accounts for almost all South Africa's infections. A key factor in the disease's spread here is the country's large migrant workforce. Epidemiologists have evidence, for example, that migrant gold miners are the link between high infection rates in the mining town of Carletonville, near Johannesburg, and sharp increases in infection rates in KwaZulu-Natal across the country. Preliminary data from studies by epidemiologist Lurie and colleagues show that HIV has infected one or both partners in 33.7% of “migrant couples” in KwaZulu-Natal—nearly double the prevalence in nonmigrant couples. In Carletonville, according to epidemiologist Brian Williams, who works with the nonprofit Council for Scientific and Industrial Research, he and his co-workers have found that a staggering 60% of 25-year-old women are infected, as are 50% of 32-year-old men (see right-hand graph). Says Williams: “It's completely terrifying.” ## Learning to govern In response to this mounting crisis, the government has organized AIDS counseling and training centers, promoted the use of condoms, led the international campaign to pressure pharmaceutical companies to lower drug prices, and boosted the AIDS research budget. But as MRC president Malegapuru William Makgoba pointed out in a 19 May Science editorial (p. 1171), the best chance to control the HIV epidemic in South Africa was when the apartheid era was ending in the early 1990s, but at the time the country “had no effective government.” Makgoba wrote: “In the midst of the heroic efforts to build a new, pluralistic South Africa, the HIV epidemic simply became one challenge too many.” Compounding the problem, the postapartheid government that took over in 1994 and which Mbeki now heads has sown confusion and anger with its HIV/AIDS policies. “You have a government that's learning to govern,” says Karim. “It's a government learning that liberation is more than good ideas.” First came the scandal over Sarafina II, a 1995 AIDS musical commissioned by the Health Department that cost$14 million rand (US\$2 million today)—a substantial chunk of the AIDS budget—and triggered an official investigation of then-health minister Nkosazana Zuma and others. Zuma, with support from then-deputy president Mbeki, next helped promote Virodene, a supposed anti-HIV wonder drug that turned out to be worthless. A team appointed by Zuma then recommended abolishing the government-appointed Medicines Control Council, which had criticized the testing of Virodene. The council, the equivalent of the U.S. Food and Drug Administration, also has written two reports detailing how AZT benefits outweigh risks, but the Health Ministry has not accepted either report.

Since taking office in June 1999, Mbeki has further baffled AIDS researchers and advocates. Under pressure to provide AZT to HIV-infected pregnant women and rape victims, Mbeki last October told his Parliament that doing so would be “irresponsible” because “the toxicity of this drug is such that it is in fact a danger to health.” Then Mbeki began questioning whether HIV causes AIDS (Science, 28 April, p. 590). This spring, the South African Parliament also learned that the health department had not spent 40% of its AIDS budget. “It's a very difficult situation we're in,” says Quarraisha Abdool Karim, the former head of the National AIDS Program (and Slim's wife), who now studies HIV infection in women. “Every few months we have this case that undermines all the good work that's been done.”

AIDS researchers in South Africa become most exercised about the country's refusal to provide drugs to HIV-infected pregnant women. South Africa participated in the PETRA study, which showed in February 1999 that a short course of AZT combined with the anti-HIV drug 3TC could cut infection rates at birth in half. Health Minister Mantombazana Tshabalala-Msimang, an obstetrician gynecologist, says the treatment is impractical in South Africa: “AZT/3TC is beyond what this country can afford.” She notes that nine sites in South Africa now are evaluating the much cheaper regimens of nevirapine (see p. 2160). “Researchers advise governments,” says Tshabalala-Msimang. “They're doing studies. They said we should wait.”

Quarraisha Karim bristles at this. “Research is being used as an excuse not to make policy,” she says. Makgoba adds that the MRC has presented the government with studies that prove the cost-effectiveness of providing AZT/3TC. He thinks part of the reluctance to follow the advice of scientists comes from the government's antipathy toward pharmaceutical companies. “It's an unhappy relationship,” says Makgoba. “And the science is caught in the middle of it.”

Parks Mankahlana, Mbeki's spokesperson, confirms a widespread rumor that providing treatment to infected pregnant women worries the government because of the number of surviving orphans this policy would create. “A country like ours has to deal with that,” insists Mankahlana. “That mother is going to die, and that HIV-negative child will be an orphan. That child must be brought up. Who's going to bring the child up? It's the state, the state. That's resources, you see?”

Obstetrician James McIntyre, who co-directs the Perinatal HIV Clinic at Soweto's Chris Hani Baragwanath Hospital, says this cynical argument doesn't even make sense. “Drug treatment doesn't create the orphan problem,” says McIntyre. “If we do nothing, seven out of 10 kids will be uninfected, and they'll be orphaned. And we have to consider what happens to orphans. It's a whole lot easier for the family to take care of uninfected children.”

When asked how Mbeki feels about embarrassing his country's AIDS researchers, Mankahlana responds, “People are just being foolish.” He adds: “You know, Mbeki wants to wake up in the morning and see the South African economy grow. He wants to see jobs being created. He wants to see crime levels coming to zero. He wants to see lodgings of people improve. OK, there's this problem of HIV/AIDS, but you see scientists must work. They must do what they have to do, which is to find a cure for this thing. They must. If they want government assistance, they will get it. End of story.”

Of course that's not the end of the story. When thousands of AIDS scientists and hordes of journalists flock to Durban in July for their weeklong international meeting, they will focus the world's attention on South Africa's response to HIV and AIDS. Makgoba suspects the conference may go a long way toward helping his government develop a sound strategy for confronting the epidemic. “When South Africa was tearing itself apart with apartheid, the world came to South Africa and solved the problem,” says Makgoba. “When South Africa is tearing itself apart with AIDS, it may just provide another opportunity for the world to help.”

29. PROFILE

# A Research Renaissance, South African Style

1. Jon Cohen*
1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl.

Photos by Malcolm Linton

Cape Town, South AfricaSitting in a patio restaurant at the posh Cape Grace hotel in the shadow of the majestic Table Mountain, Malegapuru William Makgoba shakes his head in disbelief at his own words. Speaking of his friend President Thabo Mbeki, who recently embraced the “dissident” faction that questions whether HIV causes AIDS, Makgoba says, “The sad part is, he's trying to politicize scientific facts, and that's what the Nazis did.”

Makgoba, the first black president of South Africa's Medical Research Council, has been among the most outspoken critics of Mbeki's waffling on HIV. Shortly after the news broke this winter that Mbeki had doubts about the link between HIV and AIDS, Makgoba launched a high-profile, frontal assault, including sharply worded editorials in leading South African newspapers and in Science. It's not a particularly comfortable position for a man who shares many of Mbeki's political views. Just this morning, for example, Makgoba will appear on a television talk show to discuss a book he edited called African Renaissance, for which Mbeki wrote the prologue.

Indeed, if somebody had told him that he would return in 1994 from a self-imposed exile to help improve the lot of black scientists, rise to prominence in a black-led administration, and then become one of the sharpest critics of that government, “I would have said they were crazy,” Makgoba says.

Raised in a part of South Africa's rural Transvaal that's now called the Northern Province, Makgoba grew up as a shepherd. “The first time I wore underpants was when I was 15,” he laughs. Before going off to boarding school, he rubbed lion fat on his body to make him strong.

Makgoba earned a medical degree in Durban's University of Natal and in 1981 won a Ph.D. fellowship at Oxford, studying immunology with Andrew McMichael. “Overall he was one of the most broadly able and interesting students I have seen,” says McMichael. “He's a good scientist, but with a mission to do something special for his people.”

After a stint at the U.S. National Institutes of Health and London's Royal Postgraduate Medical School, Makgoba returned to postapartheid South Africa in 1994 to become deputy vice chancellor at the University of Witwatersrand. Soon, he became embroiled in an ugly power struggle with 13 colleagues who accused him of gilding his resume. He fought back in the press, turning the tables on many of his accusers, and wrote a book (Mokoko, The Makgoba Affair) that won him loyal admirers and staunch critics. One detractor, a South African economist, wrote in Africa Studies Quarterly that Makgoba's recounting of his accomplishments made him sound “intellectually pompous and arrogant and utterly self-centered, if not downright egocentric.”

Makgoba laments that South Africa has no prominent black AIDS researchers. But “talking about AIDS is the tip of the iceberg,” he says: “There are just not many prominent black researchers. … It's one of our biggest challenges not just in medical research, but in the whole educational system.” Yet Makgoba himself has hung up his lab coat. “The quickest way for me to open possibilities for black people in the long term is not sitting in a lab and training them,” he says. “It's being in a position of power and impacting policy that will affect people across the board.”

As for Mbeki, Makgoba thinks his president may end up making lemonade from the lemon. “I do believe Mbeki is flexible enough that at the end of the debate he has caused, he'll make a judgment that's reasonable to us,” says Makgoba. “It may be that this case is a turning point.” It certainly has been for Malegapuru William Makgoba.

30. # Confronting Conference Complexities

1. Jon Cohen*
1. *Jon Cohen was accompanied by photographer Malcolm Linton. For a gallery of additional photos and the stories behind them, see www.sciencemag.org/feature/data/aids_africa/gallery.shl.

Durban, South AfricaFor most of his life, Hoosen “Jerry” Coovadia was a second-class citizen. As an ethnic Indian in South Africa's apartheid system, the University of Natal pediatrician had limited freedom to conduct his own research and little contact with colleagues abroad. “Access to the outside world and the research agenda was dictated by the white minority,” he says matter-of-factly. But, like that of millions of black, Indian, and “colored” citizens of South Africa, Coovadia's world changed in 1994 when South Africa held its first democratic election. Next month, Coovadia's status in the new South Africa will be unmistakable: He will chair the XIII International AIDS Conference here, the first time this high-profile gathering has been held in a developing country.

“We see this conference not only as an opportunity for scientific discourse, but as a chance to highlight South Africa—the extent of the epidemic here and what we've been able to do in this country,” says Salim Abdool Karim, who chairs the conference's scientific committee. Karim, who also is of Indian descent, notes that he and other scientists for many years supported an academic boycott imposed by international colleagues to pressure the South African government to end apartheid. “Even though it was a necessary strategy, it hurt us and the enemy,” says Karim. “Research in South Africa is still suffering from that. It's just a necessary part of the price we paid for our freedom.”

Coovadia, Karim, and the other organizers have gone to great lengths to make the conference work smoothly. They've had several meetings with representatives of pharmaceutical companies, who in years past have set up luxurious, even garish, booths to promote their latest anti-HIV drugs—which most Africans of course cannot afford. “If they keep their word, they will have much more restrained displays,” says Coovadia.

The organizers also have tried to allay, as best they can, the concerns that many foreign scientists have about visiting South Africa. Some have called for a boycott to protest South African President Thabo Mbeki's handling of the country's AIDS epidemic; others worry for their safety. Although most international AIDS conferences have issued upbeat press releases in the weeks preceding the meeting, the South Africans have put out notices arguing against a boycott and describing the various security firms they (and some pharmaceutical companies) have hired. Another press release states that a protest march against “pharmaceutical giants” will precede the opening ceremony and “will be peaceful, not ‘another Seattle,’” a reference to last year's riots that disrupted the World Trade Organization meeting in Seattle, Washington.

Coovadia says he welcomes protests as long as they don't cause harm or damage. “We ourselves grew up in an environment where we protested apartheid,” he says. Although he suspects attendance may be lower than the last few meetings—both Geneva in 1998 and Vancouver in 1996 had about 10,000 attendees—he says he expects at least 8000.

Karim stresses that South Africa offers something never seen before at an international AIDS conference: an upclose look at the AIDS epidemic, ground zero. “It's very nice to go to Vancouver. It's very nice to go to Geneva,” says Karim. “But we want the world to see what the epidemic can do to destroy a country and a continent.”