# News this Week

Science  19 Jun 1998:
Vol. 280, Issue 5371, pp. 1833
1. STRUCTURAL BIOLOGY

# Revealing HIV's T Cell Passkey

1. Michael Balter

CRYSTALLOGRAPHERS HAVE GIVEN AIDS RESEARCH A BOOST WITH THE LONG-SOUGHT STRUCTURE OF THE PROTEIN HIV USES TO ACCESS T LYMPHOCYTES. A SPECIAL ISSUE ON AIDS RESEARCH FOLLOWS ON PAGE 1855

In any war, it is vital to know your enemy. During the battle against AIDS, researchers have gathered as much intelligence as possible about HIV, the virus that causes the disease, seeking out weaknesses in its defenses that could be breached with antiviral drugs or a vaccine. One key target that researchers have put under intense surveillance is an HIV protein called gp120, which studs the virus's outer coat. By latching onto receptors on the surface of T lymphocytes, the immune cells that are HIV's primary target, gp120 allows the virus to enter the cells and reproduce. But intelligence about gp120 has been limited. Researchers have had to use indirect techniques to study the protein, which provided only a blurry picture of gp120's structure—not enough to design well-targeted drugs or mount an effective vaccine strategy. Now, however, a powerful searchlight has been focused on gp120.

In this week's issue of Nature, a team led by x-ray crystallographer Wayne Hendrickson at Columbia University in New York City and molecular virologist Joseph Sodroski at the Dana-Farber Cancer Institute in Boston reports that it has determined, for the first time, gp120's atomic structure to a resolution of 2.5 angstroms. This achievement has brought the protein into sharper focus than ever before. Moreover, as evidenced by two companion papers from the Hendrickson-Sodroski partnership—one in Nature and one on page 1949 of this issue of Science—the new closeup view of gp120 is already helping researchers understand why the antibodies that HIV-infected patients produce against the virus fail to knock it out, as well as revealing weak spots on the protein where a drug or an antibody might be able to slip in and gum up its molecular works.

“This is a big deal,” says Nobel laureate David Baltimore, president of the California Institute of Technology and head of a U.S. government advisory panel on AIDS vaccines. Immunovirologist John Moore of the Aaron Diamond AIDS Research Center in New York City agrees, calling the new findings “a major advance in our knowledge of the virology and immunology of HIV infection.” Such superlatives are all the more deserved, researchers say, because the structural peculiarities of gp120 had, over the past decade, frustrated the efforts of several competing labs to crystallize it—a prerequisite to determining its structure by x-ray crystallography. “A lot of people thought that to tackle gp120 would not be possible with current technology,” says Quentin Sattentau, an immunologist at the Center for Immunology in Marseilles, France. “Finding the structure was an extraordinary feat of perseverance.”

Much of the tough job of persevering with gp120 fell to crystallographer Peter Kwong, who joined Hendrickson's lab in 1987 as a graduate student. Kwong's first project was to team up with other colleagues to crystallize the protein CD4, which HIV uses as its primary receptor to gain entry into cells. In 1990, the group determined the structure of those segments, or domains, of CD4 that bind to gp120, an accomplishment simultaneously published by Stephen Harrison's group at Harvard University. But the next step, crystallizing a complex of CD4 and gp120 bound together, which could provide valuable information on how these two molecules interact, proved much more difficult.

Certain of gp120's features, including carbohydrate groups on its outer surface and its so-called variable (V) loops, make it flexible and irregular in shape, and hence nearly impossible to form into a crystal. “It was a tough project, and we were really stuck,” Kwong told Science. But before long, help was on its way. This came from Sodroski's group, which, like other research teams, had created mutant versions of gp120 as a way of probing which parts of the molecule are most critical to its role of leading the viral penetration of T cells (Science, 25 October 1996, p. 502). One day in 1993, Sodroski called Hendrickson to ask if his lab would be interested in trying to crystallize some of the altered gp120 molecules that Sodroski's lab was producing.

Over the next several years, Sodroski and Richard Wyatt, another molecular virologist at Dana-Farber, kept in constant touch with Hendrickson and Kwong, discussing strategy and designing one gp120 variant after another. The team was also aided by immunologists James Robinson at the Tulane University Medical Center in New Orleans and Raymond Sweet at SmithKline Beecham Pharmaceuticals in King of Prussia, Pennsylvania, who provided antibodies and CD4 molecules.

Kwong put the gp120s together in various combinations with CD4 and antibodies, looking for a molecular complex that was rigid enough and regular enough in shape to crystallize. Finally, he hit on the right combination: When he mixed the gp120-binding domains of CD4 with a truncated version of gp120—missing most of its carbohydrate groups, some of its V loops, and a few other bits—and then added in a fragment of an antibody called 17b that only binds tightly to gp120 when it has already bound CD4, this complex produced crystals up to 50 micrometers across. This size was just adequate for x-ray crystallography. Because the complex contained the regions of both gp120 and CD4 that previous studies had shown were critical for HIV entry into cells, the crystal structure was guaranteed to contain a gold mine of information.

“For the last 10 years, crystallographers have wanted to get this structure,” says Baltimore. “It took deep insight to know how to shave down the molecule to get a core that would crystallize.” Hendrickson adds that he is reasonably confident that the structure of the gp120-CD4-17b complex, despite its abbreviated form, accurately reflects the relationships among its three partners. He notes, for example, that the strength of the binding between the CD4 domains and the gp120 core is “quite similar” to that of the native proteins. Moore, too, says that he has “no concerns about the modifications necessary to crystallize this molecule,” adding that “whatever minor limitations exist are massively outweighed by what has been learned.”

The structure obtained from those crystals immediately began to reveal new insights. It shows, for example, that gp120 is divided into two principal domains, which are connected by four polypeptide strands that form a “bridging sheet” between them. Moreover, although most of the carbohydrate groups and V loops have been removed, these parts of the protein—whose positions can be deduced from the remaining core structure—appear to partly cover and obscure the binding sites for both CD4 and another receptor, called CCR5, that HIV-1 must also attach to in order to infect cells. This information, elaborated further in the companion paper in Nature, helps explain why HIV-infected patients produce so few antibodies that are effective in blocking these regions. The immune system essentially cannot “see” these prime targets and generate a response to them.

Another obstacle to immune attack is revealed by the detailed structure of the CD4-binding site itself. CD4 fits into a recessed pocket in gp120, which, the authors say, may simply be too deep to be easily accessed by antibodies. Moreover, the surfaces of gp120 and CD4 make only partial contact, leaving two relatively large cavities between the molecules. One of these cavities is filled with water molecules, which may provide another mechanism for HIV to escape antibody attack: Because the region of gp120 facing this cavity does not make intimate contact with CD4, the amino acids that make it up are freer to undergo mutation and avoid recognition by antibodies. “The mechanisms used by gp120 to resist [the immune system] are nicely revealed” by the new structure, says Moore.

An additional key insight gleaned from the structure, described in the Science paper, is that the bridging sheet—which is partially bound by the 17b antibody—also helps make up part of gp120's binding site to CCR5. This finding solves a mystery about how HIV recognizes CCR5, whose normal function is to bind immune regulatory molecules called chemokines. The virus had been thought to attach to CCR5 primarily through one of the V loops, called V3. But the amino acid sequence of V3 varies greatly from one HIV strain to another. In fact, it mutates so readily that the changes enable HIV to switch, after the early stages of infection, from using CCR5 to infect cells to using another chemokine receptor called CXCR4 in later stages. But that high degree of variability caused researchers to wonder how all the different variants could bind so tightly and specifically to the genetically conserved chemokine receptors in the first place.

Sodroski and Wyatt provide the answer. Using the crystal structure as a guide, they designed a series of gp120 mutants that contained alterations in the polypeptides implicated in this binding. They discovered that the chemokine receptor binding site on gp120 includes elements of both the highly variable V3 loop and the bridging sheet, which is highly conserved. This new finding explains how this site can have both the specificity required for tight binding to one chemokine receptor and still be variable enough to allow the virus to switch later to another.

Researchers are confident that the insights they are gleaning from the structure will open new doors to drug and vaccine intervention against HIV. For example, Sodroski points out that the conserved region of the CCR5-binding site on gp120 “gives people a [drug intervention] target that looks much more attractive” than either the variable V3 loop or the chemokine receptors themselves. Targeting the receptors, for example, could end up disrupting the immune systems of already immune-compromised patients. In contrast, blocking the binding site on the viral protein might be less likely to have harmful side effects. Other possible drug targets, team members say, are the cavities formed when CD4 binds to gp120. “That is where you really want to put in some therapeutic compound” to disrupt the binding, says Kwong.

Vaccine strategists will find much to mull over as well. For some researchers, the finding that the CD4- and CCR5-binding sites are essentially invisible to the immune system constitutes additional evidence that basing vaccines on whole gp120 molecules is doomed to failure (Science, 30 January, p. 650). “This plays right into HIV's defenses,” Moore says. “Knowing what not to attack is sometimes extremely valuable.”

As for what would be effective, Marseilles's Sattentau suggests that one way to get around gp120's defensive shroud might be to use a vaccine equivalent of a one-two punch. Thus a vaccine might be designed to stimulate antibodies in high enough concentrations that they could penetrate the shroud and attach to CD4-binding sites, possibly triggering a change in gp120's shape and exposing the CCR5-binding sites—which would then be vulnerable to attack by a second antibody. But whatever the winning strategy—or strategies—turn out to be, Sodroski says that the gp120 structure should give a big boost to vaccine research. “Over the next year or so, we are going to see a lot of people doing all kinds of modifications to gp120,” with the aim of finding ways to induce stronger antibody responses.

It may take some time before AIDS researchers figure out how to take maximum advantage of the weaknesses now revealed in gp120's structure. But they are confident that this intimate new knowledge of the protein will lead to a whole new battle plan and—with luck—new weapons that could deliver the knockout blow.

2. ASTROPHYSICS

# Neutron Stars Spin Out Gravity Waves

1. Meher Antia
1. Meher Antia is a science writer in Vancouver.

Newborn neutron stars may be powerful beacons of gravitational waves, the ripples in the fabric of space-time predicted by Einstein's theory of gravity. Calculations reported in the 1 June Physical Review Letters suggest that just after a giant star collapses to form a neutron star, the hot, superdense matter quivers in a way that sheds large amounts of energy and angular momentum into gravitational waves. In as little as a year, the process could slow a spinning neutron star from 1000 rotations a second to a leisurely 100, which may explain why astronomers have never spotted a fast-spinning young neutron star. “Why you don't see young neutron stars spinning fast has long been a puzzle,” says Kip Thorne, a Caltech astrophysicist. “Now there is a compelling explanation.”

Although no one has directly detected gravitational radiation, Einstein's theory holds that clumps of matter emit it when they shake or move, just as moving charged particles spew electromagnetic radiation. People, cars, and even Earth emit only trivial amounts of gravitational radiation. But under the right circumstances, supermassive objects like neutron stars can shed vast amounts of energy and momentum as gravitational waves.

Neutron stars should be born with plenty of angular momentum, because they form in supernovae—the explosions triggered when a giant star suddenly collapses to a tiny fraction of its radius. The progenitor star may be spinning slowly, but it revs up when it collapses, like the proverbial skater drawing in his arms. Yet the highest spin rates are seen not in newborn neutron stars but in older, cooler stars that have gained spin when a companion star dumped material on their surface.

Theorists have known for some time that neutron stars can lose spin through gravitational radiation, but they thought that the loss was minor. The sloshing that generates the gravity waves, they thought, would be damped by friction in the fluid of neutrons and electrons making up the star.

Last year, though, Nils Andersson of Washington University in St. Louis found that a type of vibration pattern, or mode, in the rotating star grew stronger instead of weaker when the star emitted gravitational radiation. Called r-modes, the vibrations are a bit like the currents in oceans, says Lee Lindblom, an astrophysicist at Caltech. Lindblom and his Caltech colleague Benjamin Owen, with Sharon Morsink, a physicist at the University of Wisconsin, Milwaukee, have now calculated the exact strength of the interaction between gravitational radiation and the r-modes. They found that in the hot fluid of a young neutron star, the feedback leads to gravitational radiation so powerful that it slams the brakes on the star's spin.

Although experimentalists are now building a huge gravitational wave detector called LIGO, it is designed to detect the much more powerful waves that might be emitted when two neutron stars collide. Thorne says that after a round of enhancements, LIGO might be able to pick up the gravitational waves from a newborn neutron star. But in the short term, prospects for detection aren't bright, says Lars Bildsten, an astrophysicist at the University of California, Berkeley: “Having said that, I find just the astrophysical implication of the rapid spin-down very exciting.”

3. ASTRONOMY

# Giant Survey Wallpapers the Sky

1. James Glanz

San Diego—A roll of photographic film 10 paces long marks the first step in a survey that will reach billions of light-years into space. The Sloan Digital Sky Survey's electronic camera, mounted on a 2.5-meter telescope in New Mexico, collected its first images of stars, galaxies, and quasars on the nights of 9 and 10 May (Science, 29 May, p. 1337). Converted into a photographic scroll, just 1% of the data from the first night drew gasps last week when five members of the team unrolled it during a press conference here at a meeting of the American Astronomical Society.

This first glimpse of the Sloan universe revealed a panoply of galaxies so varied that Michael Shull, an astrophysicist at the University of Colorado, Boulder, initially thought it could only be a computer simulation. “Oh, this is data?” asked a surprised Shull. “Right before us we're seeing the whole zoo of galactic morphology types. Here's a lovely one—an edge-on spiral,” he said as he began scanning the images more closely.

After an initial shake-out period, the Sloan—a collective effort of eight universities and other institutions—will collect images of roughly 200 million objects over 5 years. The project will also determine “redshifts,” or approximate distances, to the million brightest galaxies, revealing the large-scale structure of our corner of space. Equipped with six rows of charge-coupled devices—electronic light-gathering elements—the Sloan camera gathers the data across six strips of sky at once, in five colors, as Earth's rotation slowly pans the fixed telescope across the sky. “Not only are we wallpapering the sky, but our machine is doing six rolls at once,” said Michael Turner of the University of Chicago, the Sloan survey's spokesperson. Eight minutes of data are shown above.

There were no major problems getting the enormously complicated camera working the first time, says Constance Rockosi, a team member at the University of Chicago: “When we put it on the telescope and opened up the shutter, it worked.”

4. ASTRONOMY

# Supernova and Gamma Burst Might Have Common Source

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

San Diego—A stellar explosion in a galaxy 140 million light-years away might turn out to be the Rosetta stone for scientists trying to make sense of gamma ray bursts (GRBs). “This could well be the missing link” that astronomers have been looking for to explain these cataclysms, which occur at great distances and pump out staggering amounts of energy, says theorist Stan Woosley of the University of California, Santa Cruz. Woosley described the discovery and what it might mean here at the summer meeting of the American Astronomical Society (AAS).

The supernova, which two Dutch astronomers reported on 7 May in an International Astronomical Union circular, produced more than the usual cosmic fireworks. It also took place at about the same time and at the same point in the sky as a 25 April gamma ray burst, marking the first time that a burst has been linked with anything more specific than a bright spot in the sky. If the link between the supernova and the GRB holds up, it could favor a scenario proposed by Woosley and Bohdan Paczýnski of Princeton University, in which the collapse of a massive star—a process akin to the one that triggers many supernovae—sparks the outpouring of gamma rays.

The match-up of the supernova and the burst has impressed some other astronomers as well. Chryssa Kouveliotou of the Universities Space Research Association at NASA's Marshall Space Flight Center in Huntsville, Alabama, says it is unlikely to be a chance alignment, but like most astronomers, she's not fully convinced. Ralph Wijers of the State University of New York, Stony Brook, is more skeptical, saying, “the connection [between the supernova and the GRB] is unclear.”

Over the last year, the search for a GRB mechanism has heated up as astronomers have realized that many bursts occur in the far reaches of the observable universe, implying that they are unimaginably powerful. Indeed, a burst on 14 December of last year was so far off that it may have been the mightiest explosion ever glimpsed, aside from the big bang itself (Science, 24 April, p. 514). In one scenario, the bursts are triggered when two neutron stars (or a neutron star and a black hole) collide and merge. Woosley and Paczýnski's theory instead attributes them to the sudden collapse of a very massive star into a black hole, a mechanism that Paczýnski has dubbed a “hypernova.”

The 25 April burst could tip the scale toward the hypernova or, as Woosley prefers, “collapsar” mechanism. Both the Italian-Dutch BeppoSAX satellite and NASA's Compton Gamma Ray Observatory picked up the burst, and one of BeppoSAX's wide-field x-ray cameras, built by the Utrecht laboratory of SRON (Space Research Organization Netherlands), was able to pinpoint its position. Titus Galama and Paul Vreeswijk, both from the University of Amsterdam, pointed an optical telescope at the European Southern Observatory at La Silla in Chile at the spot—and found a supernova at the same position.

The supernova, designated 1998bw, had flared up in a galaxy at a modest distance of 140 million light-years. Galama and Vreeswijk, who describe their discovery in a paper submitted to Nature, declined to comment on the exact timing of the supernova with respect to the burst, but according to Woosley, they coincided “within a few days.” He says the chance that they are unrelated is only one in 100,000 or one in a million.

The 1998bw event appears to have been no ordinary supernova. Prompted by Galama and Vreeswijk's discovery, Mark Wieringa and his colleagues at the Australia Telescope National Facility observed it with a radio telescope and found that it far outshone other supernovae at radio wavelengths. Woosley adds that its visible-light spectrum, too, was “unlike that of any other supernova”—notably lacking the spectral lines of ionized hydrogen and helium.

The unusual appearance of the supernova suggests to Woosley and other theorists that it had an unusual mechanism, which could have sparked the gamma ray burst as well. The stellar implosion that triggers many supernovae generally stops when the core of the star has collapsed into a superdense neutron star; a shock wave, supercharged by a burst of neutrinos from the core, then blows off the outer layers of the star. But in a collapsar, the progenitor star is so massive—more than 25 times the mass of the sun—that the collapse would not stop at a neutron star; instead, it continues all the way to a black hole, which quickly swallows up almost all of the star's mass.

According to Paczýnski and Woosley, the matter vanishing down the gullet of the black hole emits a titanic surge of energy—the gamma ray burst—and triggers a shock wave. The shock wave heats the sparse matter near the star, creating a supernovalike display—but with telltale differences, like those in 1998bw. Wijers, however, doubts that this mechanism can explain all GRBs, saying, “we would've seen earlier associations between gamma ray bursts and bright supernovae.”

If the 25 April burst did come from 1998bw, a relatively nearby supernova, the gamma ray flash must have been 100,000 times less energetic than the burst of 14 December, which astronomers have nicknamed “Big Bang 2.” To some astronomers, that implies that GRBs may have several different mechanisms. But Woosley thinks the collapsar model can explain such widely varying events. A magnetic field rooted in the rotating black hole could extract extra energy and deliver it to nearby matter to boost some events, or a torus of material around the equator of the black hole could channel energy in two directions, making a very distant burst look brighter if one happened to be aimed toward Earth.

Less direct support for the collapsar idea comes from hints that many GRBs occur in dusty, star-forming regions of galaxies. As Marc Metzger and George Djorgovski of the California Institute of Technology in Pasadena described at the AAS meeting, several bursts have actually been traced to distant star-forming galaxies, while others have very red optical afterglows, a sign that the light has traveled through dust. “Dust also might explain the complete absence of optical counterparts in other cases,” says Djorgovski. Because the most massive stars—the ones that could spawn collapsars—live only for a few million years, they would be expected to die in the same star-forming regions in which they were born.

No one is convinced that supernova 1998bw has provided the answer to the GRB enigma. But in a year when the clues to the puzzle have come thick and fast, astronomers think their luck may be holding.

5. PALEONTOLOGY

# Parsing the Trilobites' Rise and Fall

1. Robert Irion
1. Robert Irion is a science writer in Santa Cruz, California.

In Earth's first evolutionary flowering, the Cambrian explosion, the trilobites—hardy, shelled arthropods beloved of fossil hunters everywhere—were big winners, quickly filling many niches on the ocean bottom. But they seemed to miss out on the next evolutionary burst, about 475 million years ago during the Ordovician period. While mollusks, corals, and stationary filter feeders diversified rapidly and laid the foundation for today's sea-floor ecosystem, trilobites seemingly slid into a protracted decline. By 250 million years ago, they had disappeared for good.

Now on page 1922 of this issue, paleontologist Jonathan Adrain of London's Natural History Museum and two colleagues report that this widely accepted picture of the trilobites' fate is too simplistic. With a detailed new survey, they show that while one group, or fauna, of trilobites faded precipitously during the Ordovician, the other thrived. A still-mysterious combination of ecology, geographic distribution, and high rates of speciation evidently gave this fauna an evolutionary edge.

Other paleontologists are impressed. “The contrast between the two faunas is amazing,” says David Jablonski of the University of Chicago. “You really get the sense of trilobites as a boom-and-bust group.” Adds Arnold Miller of the University of Cincinnati: “This is an unassailable data set.” But paleontologist Jack Sepkoski of the University of Chicago notes that even the more rapidly diversifying trilobites were laggards compared with other Ordovician organisms.

The new view of trilobite diversity stems from a reanalysis of the literature, led by Adrain. The team identified 945 genera and grouped them into 56 families that share common features, such as unique shell segments and shapes. They were then able to realize for the first time that the trilobites cluster into two major groups of families. Members of one cluster, the Ibex Fauna, dominated the start of the Ordovician but then grew less diverse and vanished at the end-Ordovician mass extinction 440 million years ago. The other cluster, the Whiterock Fauna, tripled their genera in the Ordovician and skimmed through that extinction virtually unscathed; not until 30 million years later did they start to wane.

These differences in rates of evolution are “the most compelling clue to [the faunas'] strikingly disjunct fates,” the authors say. Although the Ibex Fauna did not stray far from the Cambrian forms, the Whiterock Fauna rapidly evolved novel shapes and spread into new niches, which may have insulated them from extinction. “Something different about this extinction allowed the more diverse families to survive,” says Adrain, noting that diverse families weren't as protected in most other mass extinctions.

Ecology and geography may also have helped dictate the faunas' fates. Whiterock species preferred middepth environments in the ocean, while their Ibex cousins lived either in shallower or deeper water. And although tropical trilobites, including the Whiterockians, fared well, most high-latitude trilobites perished. No one knows exactly why these differences were important, but the extinction probably involved glaciation—which could have different effects at different latitudes—and the reorganization of ocean currents, adding up to what co-author Richard Fortey of the Natural History Museum calls a “global oceanic crisis.”

To disentangle these factors, “we have to get out of the library and back into the field,” says co-author Stephen Westrop of the University of Oklahoma, Norman. New fossil finds and studies of ocean conditions may reveal why some trilobites survived while others scuttled into silence on the sea floor.

6. EVOLUTIONARY BIOLOGY

# Females Pick Good Genes in Frogs, Flies

1. Elizabeth Pennisi

Finicky females have long mystified both suitors and evolutionary biologists, particularly when the female tends to pick the most flamboyant male, even if he doesn't appear to have any other redeeming qualities. Almost 70 years ago, R. A. Fisher threw up his hands and suggested that such preferences as the peahen's yen for a showy tail are arbitrary—whims that set off an evolutionary race won by the most outlandish male. By the mid-1970s, however, some biologists argued that females are not only finicky but wise. The exaggerated traits, they theorized, are a sign of less obvious “good genes” that will lead to fitter offspring. But testing these ideas has been difficult, as many factors can influence the success of offspring. “There's been a dearth of [good] data,” says Richard Howard, a behavioral ecologist at Purdue University in West Lafayette, Indiana.

Now, on page 1928, Allison Welch and her colleagues at the University of Missouri, Columbia, present an elegant series of experiments that demonstrate good genes at work. They report that male tree frogs with long calls—known to be favored by females—sire higher quality young than those with short calls. The work fits well with a handful of other studies analyzing good genes, including a study in stalk-eyed flies that links the long stalks preferred by females to an unusual genetic advantage in males.

Together, these and other studies have convinced skeptics that finicky females are actually choosing good genes, although researchers disagree on whether the effect accounts for most female preferences. “I think the good-genes theory is coming into its own,” says Mike Ritchie, an evolutionary geneticist at the University of St. Andrews in Scotland.

Few female tastes seem more whimsical than that of the gray tree frog (Hyla versicolor). Male frogs attract mates with calls that last from half a second to 2 seconds per call. Behavioral biologist H. Carl Gerhardt from the University of Missouri, Columbia, found that females head for long calls heard through a loudspeaker, even if the short calls are closer and louder. “They avoid the very short calls,” he says.

To explore whether the longer call signaled good genes, Welch, a graduate student, manipulated frog reproduction. For 2 years she removed the eggs from about 10 female tree frogs, then fertilized half the eggs with sperm from a short-calling male and half with sperm from a long caller. Next, working with Missouri ecologist Raymond Semlitsch, she compared how the offspring fared as tadpoles and after they metamorphosed into frogs, measuring their growth rates under regimes of scarce and plentiful food. Descendants of long callers won out. “Every single significant effect was in favor of the long callers,” says Gerhardt.

Other researchers praise the study because it neatly circumvents problems plaguing other “good genes” experiments, such as biased provisioning of eggs by the mother, which alters the offspring's chances. And unlike many studies, Welch's work can rule out the possibility that flamboyant males offer some benefit to their offspring other than good genes, such as food or rich territories, says population geneticist Mark Kirkpatrick of the University of Texas, Austin. Male tree frogs have no contact with their offspring except for fertilization, so their only contribution is genetic. “It's one of the most convincing documentations [of the “good genes” theory] that I've seen,” says Brian Charlesworth, a population geneticist at the University of Edinburgh in Scotland.

Kirkpatrick adds that Welch's large sample sizes will help the researchers quantify the extra fitness that long callers confer on their offspring: The study “gives us some numbers that will give us an estimate of how strong the ‘good gene’ influence is [in these organisms].” But Welch's team says it may be a while before they can answer the biggest question of all—what makes the long callers and their descendants more fit.

Another recent experiment, however, may have identified an actual genetic advantage that accompanies the females' seemingly arbitrary tastes. Gerald Wilkinson, a behavioral ecologist at the University of Maryland, College Park, and his colleagues studied species of stalk-eyed flies in which females generally outnumber males. The team determined that the biased sex ratio was caused by certain “selfish” genes on the X chromosome, which somehow attack Y-bearing sperm. As a result, during mating, the male contributes many X-bearing sperm but few Y's, and the next generation has more females than males. (In flies, as in humans, males have one X and one Y chromosome, while females have two X's.) These selfish genes are overly represented in each successive generation.

But some males have a gene on their Y chromosome that protects against the “selfish” X. And in the 15 January issue of Nature, the team showed in selective breeding experiments that this protective gene is linked to longer eyestalks. Populations descended from long-stalked males had an even sex ratio, and sometimes more males than females, showing that they had overcome the effects of the selfish X chromosome.

The researchers surmise—although they did not prove—that female flies prefer long stalks because they are genetically linked to a gene that blocks the selfish X chromosome, allowing the birth of males. That helps propagate the female's genes because their sons, as scarce males among females, will likely have many offspring. The work “provides a powerful but unusual example of how the ‘good genes’ mechanism can operate,” says Kirkpatrick, who was once something of a “good genes” skeptic. The Welch experiment, in contrast, shows the expected and perhaps more general outcome of good genes—enhanced fitness in the offspring.

For frogs, longer calls are costly to sustain, suggesting that the same genes—perhaps for stamina or more efficient metabolism—may underlie both calling and the enhanced survival of the young. But in the flies, the only link between long stalks and the restored sex ratio may be that the genes for both traits lie near each other on the Y chromosome. In either case, the female's choice is far from capricious, says Gerhardt. Instead, “they are making [life] better for their offspring”—and that's what good genes are all about.

7. DEVELOPMENTAL BIOLOGY

# Embryo's Organizational Chart Redrawn

1. Gretchen Vogel

Organizing the developing nervous system requires a formidable bureaucracy. In the classic organizational chart, the notochord, a rodlike group of cells running from the embryonic head to the tail, is the CEO of neural cell fate. Early in development, it sends out a command to tissue destined to be part of the spinal cord, ordering it to differentiate into a specialized sheet of cells called the floor plate. In turn, the floor plate—the middle manager in this hierarchy—sends out signals that trigger the formation of motor neurons, which transmit signals between the spinal cord and muscles.

But now both developmental and genetic evidence suggests that this textbook version of the organizational chart is wrong. Work presented last month at a meeting of developmental scientists* by Nicole Le Douarin of the College de France in Nogent-sur-Marne and her colleagues suggests that the floor plate is not ordered into existence by the notochord. Rather, it is another member of the executive committee, forming at the same time and from the same group of precursor cells as the notochord itself.

This and genetic studies in the zebrafish are forcing embryologists to reconsider some of their basic assumptions. “It really does change the view of how the patterning of the nervous system happens,” says developmental biologist Igor Dawid of the National Institute of Child Health and Human Development.

The view that the floor plate is created in response to orders from the notochord grew out of a set of elegant experiments begun in the 1980s. Biologists had already shown that early in development, a sheet of cells called the neural plate curls into a tube that eventually develops into the spinal cord. In the middle of the ventral side of the tube is a layer of cells called the floor plate. If the floor plate fails to form correctly, an embryo's nervous system goes haywire, with some neurons failing to develop and others becoming misdirected. Because the floor plate sits next to the notochord, known to influence various events in development, researchers tested its effects on floor-plate formation.

By either removing bits of notochord from chick embryos or grafting in extra pieces, researchers found that with extra notochord tissue, a second floor plate appears; without part of the notochord, part of the floor plate was missing (Science, 16 November 1990, p. 985, and 15 November 1996, p. 1115). Scientists were even able to identify the notochord's molecular messenger: a powerful protein called Sonic hedgehog, which in culture can cause certain cells to develop characteristics of the floor plate.

But Le Douarin's work indicates that this classic hierarchy is too simple. In very early chick embryos, her team replaced a group of notochord precursor cells with the corresponding cells from a quail embryo. Chick and quail cells are readily distinguishable, making it easy to track the fate of the transplanted cells. A few days later, the team found that the resulting chimera had formed a notochord and a floor plate—both made of quail cells. If the notochord were directing the floor plate to form, says Le Douarin, the floor plate would have been of chick cells.

Genetic data also support the floor plate's independence from the notochord's signal. Zebrafish mutants missing a notochord can develop floor plates, and distinct notochord and floor plate cells can be detected in very early zebrafish embryos, says developmental geneticist Marnie Halpern of the Carnegie Institution of Washington.

To Le Douarin, these results mean that the precursors of both the notochord and the floor plate are present in the very early embryo and move together through the developing neural tube. She suggests that previous experiments may have been misinterpreted, and that when the notochord was removed, the floor plate predecessors came along.

But Columbia University developmental biologist Tom Jessell, who did some of the earlier experiments, disagrees. He says previous work shows that the floor plate may be a team project, of notochord precursor and other cells, but the commands still originate from the notochord. The chimera experiments don't indicate when the floor-plate cells begin to differentiate, he says. He argues that the notochord's production of Sonic hedgehog is crucial, noting that in Sonic hedgehog knockout experiments, the notochord degenerates and a floor plate does not form.

Le Douarin is not surprised by the reluctance to revise what seemed an orderly chain of command. “Everybody assumed that they knew what was right,” she says, “but all assumptions are subject to revision.”

• * Molecular Genetics of Development, 6–9 May, Airlie, Virginia. Sponsored by the National Institute of Child Health and Human Development.

8. PARTICLE PHYSICS

# Heavy News on Solar Neutrinos

1. Dennis Normile

Takayama, Japan—The recent announcement by a team of Japanese and American physicists that they have found evidence of mass in neutrinos from the atmosphere stole the show at an international conference here (Science, 12 June, p. 1689). Almost lost in the excitement was a hint that the same may be true for another kind of neutrino also being captured by the thousands at the Super-Kamiokande collaboration's 50,000-ton water-filled detector, which is located in a mine 1 kilometer underground. A report of a deviation in the expected number of neutrinos generated by the sun at certain energy levels suggests that, like the atmospheric neutrinos, solar neutrinos can change their identity while in flight. That feat would require these ephemeral subatomic particles to have at least a smidgen of mass.

“We don't want to say anything too strongly about [the evidence] yet,” says Yoichiro Suzuki of the University of Tokyo, who heads the collaboration's solar neutrino analysis team. “We need more data.” But to some onlookers, the small variations in solar neutrino counts by energy reported by the team look like the signature of neutrino oscillations, or identity shifting. “It is the first time the effect has been observed within one experiment,” says John Bahcall, a neutrino expert at the Institute for Advanced Study in Princeton, New Jersey. Previous attempts have cobbled together results from different experiments, he explains.

The first clues that neutrinos might have mass came from experiments in the 1960s that picked up only between one-third and one-half the expected number of neutrinos streaming from the sun. Because neutrinos come in three types, or flavors, and the detectors were sensitive only to a subset of flavors and energy ranges, some researchers speculated that the missing neutrinos had converted, or oscillated, to a type that could not be detected. But the laws of quantum mechanics state that a particle must have mass to oscillate. That suggested a flaw in the Standard Model, the established theory of particles and forces that has served as the basis of modern physics, which posits zero mass for neutrinos.

There are at least two theories to explain where the oscillations are taking place. In one, dubbed the “just-so” scenario, solar neutrinos oscillate in a vacuum. In that picture, the number that change into an undetectable form should depend on the neutrinos' energy and the distance they travel from the sun to the detector on Earth. The theory gets its name from the fact that calculations of where particle oscillation is likely to peak required that the oscillation length be nearly equal to Earth's orbital radius. In the other, called MSW for the initials of its three inventors, the oscillations get a kick whenever the neutrinos interact with matter, whether in the sun or on Earth.

In the MSW scenario, the number of neutrinos detected should fluctuate from day to night, when Earth's rotation places the planet's entire mass between the sun and the detector. The effect should be largest for neutrinos that travel straight through Earth's dense core. The just-so picture predicts no day-night effect, but it does predict a slight variation in neutrino counts by energy and by season, as Earth's distance from the sun varies.

Earlier neutrino detectors couldn't gather enough data to test these predictions. But since it started operating in April 1996, Super-Kamiokande has recorded evidence of some 7000 electron neutrinos, more than three times the number spotted by every other neutrino detector combined.

So far, researchers have not found any day-night effect, and there appears to be no enhancement of the oscillation effect for neutrinos coming through Earth's core. The new results don't doom the MSW theory, however: Under certain parameters, the effect might not be seen for neutrinos flowing through Earth. On the other hand, Super-Kamiokande did see a small deviation from the expected spectrum curve at particularly high energies and a very slight seasonal variation in the number of high-energy neutrinos, support for the idea that oscillations can take place in empty space.

Neutrino researchers agree that more evidence is needed to make a definitive case for solar oscillations. But Bahcall and others say that the new data from Super-Kamiokande represent “real progress” in solving a decades-old puzzle.

9. COMPUTER SCIENCE

# New Language Could Meld the Web Into a Seamless Database

1. Dana Mackenzie
1. Dana Mackenzie is a science and math writer in Santa Cruz, California.

The World Wide Web, circa 1998: A chemist wants to know how to perform a certain reaction. She does a literature search on the Web and copies down by hand the chemicals she needs. She does another search to find the vendors who sell those chemicals and jots down their names. She sends a purchase order to Company A, which pays a clerk to enter the order by hand.

The Web, circa 2000: The chemist enters some words describing the reaction. An intelligent Web agent finds it and asks her if she wants to order the reagents. She clicks “Yes,” and the Web agent takes care of the rest. A few days later, she receives her chemicals.

The key to eliminating the human intervention between the chemist and her chemicals is a new Web invention called extensible markup language. XML, developed by a group of technology companies and universities called the World Wide Web Consortium (W3C), goes beyond merely displaying data—the strength of hypertext markup language (HTML), the language that currently dominates the Web—to making it meaningful for other computers.

In HTML, for example, CaCO3 is nothing more than a set of letters on a screen. But in XML and related languages, a “tag” can identify it as the chemical symbol for a compound. On another Web page, the same tag might go with the words “calcium carbonate,” and on a third, with “Catalog Item No. 1311.” An intelligent search engine can tell from the tags that they represent the same compound. Then it can download the accompanying data on, say, chemical properties or bulk prices, in a form ready for the user's software to manipulate.

Earlier this year, the W3C approved the first version of XML, and software companies are working on XML-capable browsers. Ultimately, say Web developers, the language could open the way to “a new Internet” that would be easier to search and exploit, offer more flexible formatting of documents, and might even usher in the ballyhooed age of electronic commerce. As Tim Berners-Lee, the creator of the Web, told the Los Angeles Times, “Whereas phase one of the Web puts all the accessible information into one huge book, if you like, in phase two we will turn all the accessible data into one huge database.” For scientists, says Peter Murray-Rust of the University of Nottingham in England, a contributor to XML, “I think it will make a major difference. … For the first time we've got something capable of managing most of the information we deal with.”

The success of XML, however, will hinge on the ability of professional societies and others to agree on what kind of information they want to share and how that information is to be structured, because XML offers far more latitude to its users than HTML does. Like any markup language, HTML annotates text with certain instructions or “tags,” which a computer program, such as a Web browser, can spot and act upon. Designed to serve as a simple lingua franca for the Web, HTML has a very small set of tags, all of which (at least in early versions of HTML) refer only to the display of text. The tags <H1> and <H2>, for example, tell a browser to render the ensuing text as a main heading or a subheading.

By contrast, XML allows groups of users to define their own tags. It does this by prescribing only the syntax of a file (how tags look and how they are used), and not the semantics (what they mean). The tags can, for example, contain information about the text, rather than just about how to display it. In an online bibliography, the title and author of a book could be tagged as <TITLE> and <AUTHOR>, instead of <H1> and <H2>. When a user searched for <AUTHOR> Gates, he would find Bill Gates's The Road Ahead, without wading through spurious links to transistor gates, garden gates, or even articles about Bill Gates.

The new language is actually an adaptation of SGML (for standard generalized markup language), which was developed in the 1980s for technical communication but was vastly too complicated for wide use on the Web. In 1996, Jon Bosak of Sun Microsystems formed a working group in the W3C to simplify SGML for Web use. Although the committee originally envisioned the language as an “SGML Lite,” the members eventually realized it was different enough to deserve its own name. Companies that develop Web browsers are now implementing XML in their products. Microsoft's Internet Explorer, for example, already contains an XML parser—a program that separates tags from text—as does the publicly available source code for Netscape Communicator. Parsers make it possible to transfer information from one database to another, but they can't display an XML file on screen. The W3C is now drawing up a format for the additional applications needed to display XML files: “style sheets,” which turn tags into formatting instructions.

Of course, no style sheet would be able to handle the Babel of information that would result if users all made up their own tags. For XML to work, common-interest groups will have to agree on a shared vocabulary. “The technology is the small part,” says Dan Connolly of the W3C, one of the designers of the new language. “The large part is getting people together to agree.”

That has already happened in a few cases. Mathematicians, who have long been hamstrung by HTML's inability to handle complex equations, can now post their pages in MathML, a rather straightforward application developed by the W3C in which the XML tags represent mathematical formatting instructions. A more ambitious XML offshoot called chemical markup language (CML) is the brainchild of Murray-Rust, who is a crystallographer as well as a Web expert. CML enables the viewer's computer to meld information stored in separate databases into a seamlessly linked, interactive document, made to order. By clicking on different parts of a CML page on a certain protein, for example, a user can call up windows showing its molecular structure, its sequence, and the structure of a ligand it attaches to. A click on the sequence in one window then lights up the corresponding part of the molecular model in another.

Meanwhile, computer scientists are working on the smart browsers and souped-up search engines that will make the fullest use of XML. Such “intelligent agents” would be able to answer queries current search engines can't touch: “Is there a university in a state bordering Virginia with an ROTC program, Japanese classes, and a Computational Biology major?” The answer (University of Maryland) happens to be where one of the first such search engines resides. The experimental browser, developed by computer scientist James Hendler, can make the necessary connections because it works with an advanced markup language in which the XML tags indicate not only meanings but relationships between entities (universities are located in states, and majors are found at universities).

Even XML aficionados don't expect to see these kinds of tags popping up on every Web site. For displaying ordinary text documents, HTML is likely to remain the standard, and XML-capable browsers will still be able to read pages written in HTML. But they believe that for specialized Web applications—in science, for example—XML will quickly make converts. “People predicted the Web would fail because no one would want to learn HTML,” says Tim Finin, a computer scientist at the University of Maryland, Baltimore County. The pessimists were wrong, he notes, and “the same thing will happen with XML.”

10. GENOMICS RESEARCH

# Year of the Cat--in More Ways Than One

1. Ken Garber
1. Ken Garber is a science and health writer in Ann Arbor, Michigan.

Far from the limelight shining on the human and mouse genome projects, researchers have also been laboring on the genomes of a half-dozen other mammals. One of these efforts, the Feline Genome Project, is about to hit a major milestone. Researchers at the National Cancer Institute's (NCI's) Laboratory of Genomic Diversity in Frederick, Maryland, expect to complete a genetic map this year—appropriately, the Chinese Year of the Cat.

The immediate goal of the NCI effort, which will cost about $3 million, sounds modest: a map with about 950 markers spread across the roughly 3 billion bases in the cat genome. That's far less detailed than the human and mouse maps, each with more than 20,000 markers on genomes that also contain about 3 billion bases. And there are no plans to undertake the massive job of determining the complete cat genome sequence. But the cat map could nevertheless turn out to be a useful guide to human genetic diseases. Cats and humans share almost 60 inherited diseases, including polycystic kidney disease, diabetes, heart muscle disorders, and certain common immune cell cancers. Once the cat map is in hand, the NCI group plans to use it to track down the cat disease genes and then mine the comparable regions of the human and mouse genomes for candidates for human disease genes. If the same genes turn out to be at fault in both species, then cats would also provide good models for the human diseases. Animals like cats and dogs offer a “tremendous, rich resource of genetic diseases that can't be studied in mice,” says Don Patterson, director of the Center for Comparative Medical Genetics at the University of Pennsylvania School of Veterinary Medicine. Stephen O'Brien, chief of the NCI team, began mapping the cat genome 20 years ago because he thought that it might help him find cat genes that regulate the effects of a cancer-causing feline virus. When his first map attempt (Science, 16 April 1982, p. 257) revealed that gene arrangements on human and cat chromosomes are very similar—much more so than those on human and mouse chromosomes—O'Brien quickly grasped the potential for “comparative mapping” between humans and cats. He has been working to complete the feline map ever since. For the bulk of the mapping effort, the O'Brien group is using standard crossbreeding experiments—an effort aided by the availability of Asian leopard cats that the NCI team borrowed from the National Zoo in Washington, D.C., and bred with domestic cats to produce a handsome cross known as the Bengal. The leopard cats “were in the right place at the right time,” notes NCI geneticist Leslie Lyons. By following the inheritance pattern of cat genes in blood samples taken from crosses through three generations, the researchers can establish their relative positions on the map, because genes that are located near one another on chromosomes tend to be inherited together. Breeding domestic cats with Asian leopard cats helped, because the genes of the two species are sufficiently different that tracking cat genes to the third-generation crosses can be readily accomplished. To fill some holes and add detail, the NCI team also applied a newer technique—radiation hybrid mapping—used for human genome mapping. NCI geneticist Bill Murphy irradiated cat cells, breaking apart their chromosomes, and then fused the cells with hamster cells. The cat chromosome fragments integrate into the hamster chromosomes in these hybrid cells, which are then tested for the presence of cat DNA markers. The nearer two markers are on a cat chromosome, the greater the likelihood that they will end up on the same fragment—and in the same cell. Together, the crossbreeding and radiation mapping techniques will enable the NCI team to create a map showing the relative positions of about 350 genes. But although genes are needed to match to the corresponding genes in other species, they don't vary enough from one individual to another to be used as markers for tracking down unknown genes. So again using blood samples from the crossbred cats, geneticist Marilyn Menotti-Raymond is mapping cat microsatellite markers, short segments of repeating DNA that are variable enough for tracking down genes for diseases and other traits. When it's ready, the cat map will include about 600 microsatellite markers. The next step will be to use the map to pin down genes that cause cat—and presumably also human—diseases. Lyons and Menotti-Raymond have already collected cat families with cancer, retinal atrophy, and polycystic kidney disease that they will use to track down human gene counterparts. Patterson expects the cat to be a better mirror of complex human diseases than lab mice, which he says are “essentially manmade organisms” that have lost many recessive traits and diseases through inbreeding. But the usefulness of cats as models of human disease remains to be shown. O'Brien also expects benefits from comparing the genetics of the 37 cat species, which range from the tiny sand cat to the majestic lion. “Less than a tenth of 1% of all mammalian species that have existed survive,” he notes. O'Brien believes the survivors' genes contain disease prevention secrets preserved by natural selection that could be identified by these studies. “We want to find out the things they have conjured up,” he says, “because we might not be as clever ourselves.” 11. # AIDS Research--1998 1. Barbara Jasny, 2. Daniel Clery Periodically in the history of the AIDS epidemic it has been worthwhile to stand back and do a reality check. What do we really know? What do we need to add in research and resources? In this special issue, we examine these questions with various News stories, Articles, Viewpoints, and personal perspectives. The News section focuses on Europe's contributions to AIDS research. With only a fraction of the resources available to their counterparts in the United States, European researchers have made, or been involved in, many key scientific breakthroughs. Despite these successes, reports Science's Paris correspondent Michael Balter, budgets for AIDS research are being squeezed in much of Europe. (For additional discussion of AIDS research in France, see Science, 16 January, p. 312.) The News also looks at how the United Nations' special program on AIDS and its head, Peter Piot, are struggling with limited resources to stem the tide of the global epidemic. (See Piot's essay on the AIDS epidemic on p. 1844.) In terms of the virus, major advances have been made in understanding, at the molecular level, how the viral envelope glycoproteins interact with host cell receptors (Wyatt and Sodroski, and see Rizzuto et al., p. 1949) and how viral accessory proteins are used to optimize production and avoid immune recognition (Emerman and Malim). As Varmus and Nathanson point out in their Editorial (p. 1815), the decreases in AIDS mortality in the United States are largely a scientific triumph of transmuting information about virus structure and replication into therapeutics. Side effects and virus resistance are clouds looming on the horizon, however, and treatment failure is a reality whose definition is much more complicated than merely finding resistant virus (Perrin and Telenti). Before resistance testing can be used clinically in designing individualized therapies, there are significant issues that must be addressed. As reviewed by Ho, eradication or long-term suppression of the virus may require immune-based strategies in addition to antiviral drugs. Pragmatically, tracking of the virus and analysis of viral reservoirs in the body will require better resources, which has led Cavert and Haase to propose the establishment of tissue banks from HIV-infected individuals. There is a general consensus that control of the epidemic will depend on the development of an effective vaccine. Increased understanding of cellular immunity is beginning to shed light on the immune responses likely to be protective (Letvin). Although a variety of vaccines are being tested in animal and human studies, the field is still searching for the best strategies. One of the features that must be factored into all strategies for dealing with HIV is that the virus is changing constantly (Korber et al.). Studies of virus variation and evolution will continue to be needed to design the next generation of weapons against HIV. For any individual, the best strategy is never to be exposed to the virus. However, promoting prevention is an extraordinarily complex issue, mired in societal norms and taboos. The reality is that prevention is possible. Phoolcharoen describes the history of the epidemic in Thailand, where a forthright campaign and a commitment through all levels of society to controlling sexual transmission have led to dramatic decreases in infection rates. Behavioral interventions have been demonstrated to be effective in the United States as well (The National Institute of Mental Health Multisite HIV Prevention Trial Group). However, there has been a dismal failure in the United States to take the politically unpalatable steps that would work (see the Research Commentary by Hein, p. 1905), and it is much too soon to exhibit complacency toward HIV. 12. EUROPEAN AIDS RESEARCH # Europe: AIDS Research on a Budget 1. Michael Balter EUROPEAN RESEARCHERS HAVE SCORED SOME MAJOR HITS IN UNRAVELLING AIDS, WITH JUST A FRACTION OF U.S. FUNDING. NOW EVEN THESE MODEST SUMS ARE STARTING TO SLIDE. Paris—In 1978, a desperately ill Portuguese taxi driver, who had previously spent time working in Angola, was diagnosed in Paris with a rare form of pneumonia caused by the lung parasite Pneumocystis carinii. The three Paris-area specialists who were called in to consult on the case—infectious disease expert Willy Rozenbaum, immunologist Jacques Leibowitch, and lung specialist Charles Mayaud—were baffled. But soon a handful of other patients, many also with African connections, began showing up with similar symptoms: an inability to control normally benign infections. And when, in 1981, the U.S. Centers for Disease Control and Prevention (CDC) in Atlanta published its first report of similar Pneumocystis cases in gay men in the United States, the French doctors quickly realized that they were confronting a frightening new disease of international proportions. In the 20 years since the first AIDS cases appeared in Europe, the epidemic has spread across the continent, leaving no country untouched. The European Centre for the Epidemiological Monitoring of AIDS, a Paris-based joint project of the World Health Organization and the European Union (EU), reported a cumulative total of more than 200,000 cases of the disease by the end of 1997. Although more than 90% of this toll is concentrated in Western Europe, the epidemic is now starting to grow in Central and Eastern Europe—Ukraine, for example, suffered a sixfold increase in AIDS incidence between 1995 and 1997. As illustrated on the following pages of this Special News Report, European scientists have been on the front lines of the battle against AIDS from the very beginning. Rozenbaum, Leibowitch, and Mayaud soon joined forces with other researchers, including virologists at the Pasteur Institute in Paris, to form the team that first isolated HIV, the virus that causes the disease. And French scientists made a number of other early breakthroughs in HIV research, as did researchers in the United Kingdom, Sweden, and the Netherlands. More recently, Italian, Belgian, and Swiss researchers have taken their places in the front ranks, helping to unveil the mysteries of how HIV gains entry into its target cells and developing potential new therapies to block the virus. The accomplishments of European scientists are all the more remarkable given the relatively modest sums their governments spend on AIDS research. The combined AIDS research budgets of the entire continent amount to only a fraction of the more than$1.7 billion spent annually in the United States. Even France, which has made far and away the largest and most consistent commitment, spends only about 2% of the U.S. total. Moreover, in the past few years the resources available to European AIDS researchers have begun to shrink, as one nation after another has cut its budget for AIDS research. In some countries, notably the United Kingdom, these cuts have resulted from the ending of specially earmarked funds for AIDS research; in others, such as Germany, they have exacerbated what many researchers believe is a long-standing failure to fund the scientific battle against the disease adequately (see p. 1857 and p. 1858).

View this table:

## Strength in numbers

These funding policies have left many European scientists feeling handicapped when it comes to competing with their American colleagues. “We do good work in Europe, but it is always 5 minutes later than the Americans,” says virologist Volker Erfle, of the Center for Environmental and Health Research near Munich. “This 5 minutes we would like to catch up on.” One way of keeping up with the competition, European scientists have found, is to work with it instead of against it—a strategy that has particularly profited researchers from Belgium, Switzerland, and Italy, whose collaborations with American colleagues have fueled many of the recent breakthroughs in identifying the essential “coreceptors” that HIV uses to gain entry to its target cells (Science, 8 May, p. 825).

Another strategy is to work together across Europe, rather than across the Atlantic. In the past few years, the EU has funded several programs designed to strengthen ties and foster collaborations among European AIDS researchers. “A new culture of European collaboration has been born,” says Jean-Louis Virelizier, a viral immunologist at the Pasteur Institute in Paris. Virelizier, who coordinates a network of French, British, and Spanish scientists investigating how the replication of HIV is regulated, adds that not long ago, “European scientists [would] ignore their immediate neighbors and were almost exclusively occupied by transatlantic collaborations.”

Despite the boost from these Europe-wide programs, the level of EU funding for AIDS research, at about $20 million for the period 1994–98, is quite modest compared to the national AIDS research budgets of many individual European countries. Moreover, most EU money goes to support travel and meeting costs rather than laboratory research. “The EU just puts a little cream on top” of national budgets, says Erfle. That leaves most European AIDS research at the mercy of national priorities. But the financial commitment from each European nation only loosely reflects the extent to which it is afflicted by the epidemic. France, for example, whose cumulative total of 47,000 AIDS cases makes it second only to Spain on the continent, spends$39 million on AIDS research annually, almost twice as much as any other single European country (Science, 16 January, p. 312). Spain, on the other hand, which lags behind much of Western Europe economically, spends about $530,000 per year on AIDS research, while the relatively wealthy Netherlands, with a third of Spain's population and less than one-tenth of its AIDS cases, has spent an average of$4 million annually—although these funds are now under serious threat (see p. 1859).

## Struggling to catch up

José Alcamí, a virologist at the 12 October Hospital in Madrid, says that Spain's explosive AIDS epidemic is the result of “a lack of preventive measures over the past 10 years.” To a large extent, Alcamí says, the failure to take firm action during the early days of the epidemic was due to the fact that most of those affected in Spain are in “marginalized” groups: 62% of AIDS patients in the country are intravenous drug users. And this negligent attitude, Alcamí adds, also made it very difficult for Spanish AIDS research to get off the ground: “During the 1980s, [Spanish health authorities] did not create an AIDS research agency in Spain, like the [U.K.] Medical Research Council's Directed Programme on AIDS or France's National Agency for AIDS Research.” As a result, there are only about a half-dozen research groups in Spain doing basic AIDS research on a full-time basis.

The news from Spain is improving, however. Although research spending remains very low, Spain has budgeted about $13 million for AIDS prevention programs this year and is making anti-HIV therapies available free to all patients. Francisco Parras, head of Spain's National AIDS Plan, told Science that these measures have already resulted in a decrease in newly reported AIDS cases, down 25% between 1996 and 1997. And the widespread use of antiviral therapies, Alcamí says, is now allowing Spain's hospital-based clinicians to participate in international drug trials, which are bringing Spanish researchers into closer contact with their international colleagues: “We are finally starting to create a more coordinated [AIDS research] network.” Although lack of research funds is chronic in poorer European countries such as Spain, Portugal, and Greece, scientists from nations that have traditionally made major commitments to AIDS research have recently suffered cutbacks and funding instability. An example is Italy, which ranks fourth in Europe in AIDS research spending. Italian scientists—whose roster has recently been boosted by several talented researchers returning after doing postdocs in the United States—have made key contributions to AIDS research, including important insights into how the body's immune system responds to HIV infection (Science, 7 July 1995, p. 24). “Italy's contribution has been overwhelmingly impressive,” says immunologist Jean Claude Gluckman of the Pitié-Salpêtrière Hospital in Paris. Despite this impressive performance, 2 years ago Italian health officials threw the nation's AIDS research program into disarray by suspending its funding as part of a major shake-up of biomedical spending (Science, 11 April 1997, p. 191). When the funds were finally restored late last year, AIDS research had suffered a cut of almost 20%. Although this meant considerable hardship for many research teams, some researchers say the effect of the shake-up was not all bad. “I do not think the cuts in AIDS research will affect the best research groups,” says Mario Clerici, an immunologist at the University of Milan. “On the other hand, there are dark clouds on the horizon for those researchers who are not doing cutting-edge science or who have used AIDS money as an excuse to finance other research projects.” More of a threat, say some Italian researchers, is the instability created by health ministry policies that require all AIDS grants to be renewed annually. “This creates a substantial waste of time and energy in rewriting grant applications,” says immunovirologist Guido Poli at the San Raffaele Scientific Institute in Milan. Italian AIDS researchers have been lobbying Italy's national health institute, the ISS, to change this policy. But microbiologist Antonio Cassone, head of the ISS's bacteriology and medical mycology department, says that these efforts will meet tough resistance from health authorities “as long as the institute follows the typical rules and bureaucracy of an Italian governmental institution.” ## Pulling together Given these ups and downs in national funding policies, a number of AIDS researchers told Science they would like the EU to step in and fund a greater proportion of direct research costs. “At this stage the EU should support more basic research, not just provide funds for traveling” to meetings, says Kai Krohn, director of the Institute of Medical Technology at Finland's University of Tampere. And Arsène Burny, a molecular biologist at the Free University of Brussels, says that meetings are useless unless scientists have new research to report: “It's fine to get people together, but you have to have something to say to each other.” Although most EU-funded AIDS programs do not directly support laboratory work, researchers involved in the handful of so-called “shared cost” programs, which do provide research funds, say these projects have helped boost collaborations that might have been too expensive to receive funding at the national level. “In general, our program has worked very well,” says Paolo Lusso, a virologist at San Raffaele and coordinator of an EU project studying the possible role of viruses other than HIV—particularly human herpesviruses—as cofactors in the progression of AIDS. The project, which began in 1996 and received a 3-year grant of nearly$1 million—a large amount for the EU—has brought together scientists from Italy, Sweden, Austria, Belgium, and Finland.

At the moment, however, it is unclear how AIDS research will fare when the 5th Framework Program—the multibillion-dollar umbrella for the EU's spending on science and technology—is finalized later this year. But as national AIDS budgets continue to shrink, international collaborations—both within Europe and around the world—may be key to keeping European scientists on the front lines. Says Krohn: “The economic policy of the EU is to beat the Japanese and the Americans, but as AIDS researchers, we must think globally.”

13. EUROPEAN AIDS RESEARCH

# U.K. Community Finds It's a Jungle Out There

1. Michael Balter

In the early days of the AIDS epidemic, British scientists were quick out of the starting gate when it came to making AIDS discoveries. In 1984, for example, Robin Weiss's group at the Institute of Cancer Research in London made the critical finding that HIV used a protein called CD4 as a receptor to enter the T lymphocyte immune cells that are its main target. This discovery (which was reported simultaneously by a French group) put British AIDS research on the map. The U.K. Medical Research Council (MRC) quickly followed up with a high-profile commitment to fighting the disease. Its AIDS Directed Programme, begun in 1987, “ring-fenced” special funds for AIDS research and attracted some of the nation's most talented scientists into the field.

This flying start was not to last, however. In 1994 the MRC terminated its special AIDS program, on the grounds that AIDS researchers no longer needed special treatment to compete for funds with scientists working on other diseases. The result, researchers told Science, has been a sharp decline not only in their funding but also in the morale of the AIDS research community. “There was a good esprit de corps throughout British AIDS research, and that has fallen by the wayside,” says Weiss. “There is no longer the same sense of community.” Moreover, at many research centers, the number of people working on AIDS has shrunk along with the funds available for research. “Three years ago we had 25 people working on HIV [with] MRC support,” says Andrew Leigh Brown, an AIDS researcher at the University of Edinburgh's Centre for HIV Research. “Now there are eight funded research posts, only three of which are supported by the MRC.”

MRC officials defend their decision to end the program. “By the early 1990s, we decided that we had a pretty large portfolio of AIDS research,” says Tony Peatfield, manager of the MRC's physiological medicine and infections board. “Once it became evident that there was a comprehensive effort, we thought we should level the playing field” in the competition for grant funds. Peatfield adds that the MRC also had to listen to researchers in other fields. “We had people saying to us, ‘Why is AIDS so special that it should get ring-fenced money, rather than asthma or malaria or transmissible spongiform encephalopathies?’”

Yet some AIDS researchers, while agreeing in principle with the idea of a level playing field, argue that the MRC has recently turned down even highly rated AIDS proposals. An example cited by many is a long-running program coordinated by Oswald Jarrett and James Neil at the University of Glasgow, which uses the cat version of the AIDS virus, FIV, as an experimental animal model to study possible vaccines against HIV. This project, which received the MRC's highest possible rating in its most recent review, nevertheless lost most of its MRC funding last year. Peatfield insists that AIDS research is not being singled out, pointing to the physiological medicine and infections board's decision not to fund another top-rated grant in another field of research.

The end of special AIDS funds from the MRC has led some researchers to seek money elsewhere, and some have been encouraged by the recent decision of the Wellcome Trust—Britain's mammoth biomedical charity—to begin funding AIDS research. But the trust's commitment, about $2.4 million for the 1996–97 fiscal year, covers only a fraction of the more than$8 million drop in MRC AIDS funding since its high point in 1994. The shortfall, researchers say, does not bode well for the future of British AIDS research. Says Rodney Phillips, an immunologist at Oxford University: “I think the next generation of researchers is being dissuaded from going into this field.”

14. EUROPEAN AIDS RESEARCH

# German Powerhouse Gives AIDS the Cold Shoulder

1. Michael Balter

If any country in Europe is well placed to make major contributions to AIDS research, it is Germany, traditionally a biomedical research powerhouse. Indeed, in the mid-1980s, the German pathologists Paul Racz and Klara Tenner-Racz, now at the Bernhard-Nocht Institute for Tropical Medicine in Hamburg, were the first to carry out detailed studies of HIV infection of the lymph nodes of AIDS patients, pioneering work that paved the way for more recent studies on how the virus destroys these immune system organs.

Despite this promising start, German AIDS researchers and colleagues in other countries told Science that Germany hasn't pulled its weight in the battle against the disease. “Given their excellent tradition in biology and medicine, the German AIDS effort has been, with few exceptions, stunningly poor,” says Simon Wain-Hobson, an AIDS researcher at the Pasteur Institute in Paris. And Volker Erfle, a virologist at the Center for Environmental and Health Research near Munich, agrees that “we have not played a leading role in AIDS research, neither in a European nor an international context.”

Germany's lackluster performance was clearly revealed in June 1997, when an international review board convened by the Ministry of Education, Science, Research, and Technology (BMBF)—the biggest single source of AIDS research funds—concluded that more than a third of BMBF-supported AIDS research projects no longer deserved to be funded. The BMBF duly slashed its HIV research budget for 1998. And although researchers are not happy with these spending cuts, some say they are nevertheless pleased to see German AIDS research subjected to international review. “In the past, many groups applied for grants when it was trendy to have an AIDS project running in the lab,” says virologist Andreas Meyerhans at the University of the Saarland in Homburg. “Scientifically, the research was often unfocused and of limited value.”

Researchers cite a number of possible reasons for Germany's disappointing contribution, including the fact that the nation spends relatively modest sums for AIDS research (see table on p. 1856). But Peter Lange, director of BMBF's health research unit, says that the ministry sometimes found it difficult to convince scientists to go into the field: “I often heard researchers say that if they went into AIDS research they would have no future, because AIDS is not a very socially accepted disease.” And Erfle comments that public support for AIDS research has never been very high in Germany. “When the projected figures [for AIDS incidence] did not come up to the worst case scenario, the interest from the public and the press went down considerably.”

For the moment, research officials are hoping that their trimmed-down AIDS effort—like a tree that has undergone a badly needed pruning—will now begin to bear fruit. Says Lange: “There is now some modest indication that the quality of research is improving.”

15. THE NETHERLANDS

1. Michael Balter

Amsterdam—Back in the 1980s, when most of the important AIDS discoveries were coming out of the United States, France, and the United Kingdom, Dutch AIDS researcher Frank Miedema had a credibility problem. “I would get up at a meeting, and people would say, ‘Who is this strange guy from Amsterdam; what does he have to tell me?’” Miedema recalls. But he soon got people's attention. Thanks to their access to two large cohorts of HIV-infected people in Amsterdam, gay men and intravenous drug users, Miedema and his colleagues accumulated enough data to convince skeptical researchers of a groundbreaking insight: In many patients, progress to AIDS is associated with HIV's evolution from strains that do minimal damage to their T lymphocyte target cells into viruses that easily kill T cells.

The Amsterdam team went on to map these differences in virulence to small changes in the amino acid sequences of proteins that make up HIV's outer coat, a finding that helped make sense of more recent discoveries that different viral strains use different cell surface receptors to latch onto T cells. And over the past 15 years, this unusual collaboration-led by public health expert Roel Coutinho of Amsterdam's Municipal Health Service, clinician Joep Lange and virologist Jaap Goudsmit at the University of Amsterdam's Academic Medical Center, and Miedema, an immunologist with the Netherlands' Red Cross Blood Transfusion Service-has made a number of other important contributions to AIDS research, including key findings on AIDS epidemiology, markers of disease progression, antiviral therapies, and the interplay between HIV and the immune system.

“They are the number one group in Europe for impact on AIDS research,” says immunovirologist Guido Poli of the San Raffaele Scientific Institute in Milan. And Anthony Fauci, director of the U.S. National Institute of Allergy and Infectious Diseases in Bethesda, Maryland, adds that the Amsterdam group has “made very significant inroads in understanding [HIV] pathogenesis. If you look at the size of their country [15 million people] and the size of its [financial] commitment, as opposed to some larger countries that have committed many more resources, their contributions really stand out.” Indeed, the Amsterdam team has accomplished so much that the Dutch health ministry—which had given the group more than $40 million over the past decade—cited its very success as the rationale for stopping the collaboration's funding after this year (Science, 19 September 1997, p. 1757). The team's own members ascribe their success to being in the right place at the right time. “When the AIDS epidemic started, it was clear from the beginning a virus was involved,” says Coutinho. “So we started discussing it with the virology group at the University of Amsterdam.” That brought in Goudsmit, who began trying to devise a blood test for the disease. Soon after, the group was joined by Lange, who was treating AIDS patients at the university's medical center, and Miedema, who was doing his doctoral research at the blood bank. When the first AIDS cases were diagnosed in the Netherlands in the early 1980s, the Amsterdam health service and the Red Cross blood bank were already running programs involving the city's gay male community. These included an effort to combat sexually transmitted diseases and a large-scale clinical trial of vaccines against hepatitis B. The team's access to these cohorts of people at high risk for HIV infection allowed it to gather valuable data on the course of the disease from the time of initial infection to the development of full-blown AIDS, usually many years later. Yet researchers who know the Amsterdam group well say that the collaboration has not always gone smoothly, particularly because each member of the team is endowed with a forceful personality and strong views about AIDS research. One Dutch scientist, who asked not to be identified, says that rivalries soon emerged over “priorities of research, distribution of money, and authorships.” Indeed, in an interview with Science, the group admitted that members often disagreed. “There were moments when I thought we couldn't go on,” says Coutinho. “Everyone is very ambitious, and everyone has certain goals.” But in the end, they realized that their success was based on their dependence on each other and on close contact with the cohorts. “Without these cohorts, we would be nowhere,” Coutinho says. “If we had split up, we all would have lost.” In recent years, as the team has gained international recognition, some members have become outspoken in broader debates in the AIDS research community as well. For example, Lange, who has been principal investigator on numerous clinical trials sponsored by the pharmaceutical industry, wrote a blistering critique in Science last year of the way many anti-HIV drug trials are conducted (Science, 25 April 1997, p. 548). Biting the hand that feeds much of his research, he condemned drug companies for their reluctance to carry out multidrug trials that include products of their competitors, even if such combinations might provide the best therapy. And Miedema has come out swinging in the fierce debate over just how HIV infection leads to the drastic loss of T cells that heralds the destruction of the immune system. In particular, he has tangled with David Ho, director of the Aaron Diamond AIDS Research Center in New York City, over Ho's theory that the immune system becomes exhausted in its battle with HIV when it can no longer produce T cells faster than they are destroyed by the virus (Science, 21 November 1997, p. 1399). Miedema, along with Hanneke Schuitemaker of the blood transfusion service and other Dutch colleagues, argues that the virus destroys the immune system by interfering with the regeneration of new T cells rather than by directly killing large numbers of mature T cells. The Amsterdam group's prominence on the international AIDS research scene makes it all the more ironic that the health ministry has decided to stop funding it. The collaboration's future is now in the hands of a committee appointed by the Dutch Medical Research Council, which is part of the Netherlands' science ministry. The committee is expected to make a recommendation this autumn on whether the project should continue, and if so, who should fund it. “We have funding until the end of this year,” says Coutinho, adding that if new money is not found before September, “we will have to stop the cohort studies.” Given the contributions of the Amsterdam group over the years, researchers say, that would be a major blow not only to AIDS research but to HIV-infected people throughout the world. 16. UNITED KINGDOM # Modest Briton Stirs Up Storm With Views on Role of CTLs 1. Michael Balter Oxford, U.K.—Andrew McMichael, head of the Molecular Immunology Group at Oxford University, seems an unlikely candidate to kick off a scientific controversy. But to many AIDS researchers, this soft-spoken Briton is best known for his provocative and much-debated ideas about how HIV, the virus that causes the disease, escapes the wrath of the immune system. McMichael's theory stems from his 2 decades of solid work on immune cells called cytotoxic T lymphocytes (CTLs), among the most formidable weapons in the immune system's arsenal. CTLs home in on microbe-infected cells and destroy them before the microbe can reproduce and infect neighboring cells. In 1986, McMichael, along with Oxford University immunologist Alain Townsend and other U.K.-based co-workers, reported crucial evidence that CTLs move in for the kill when they recognize small fragments of microbial proteins that have been transported to the cell's surface. A decade later, McMichael, along with colleagues at Stanford University, unveiled a revolutionary new technique for identifying and quantifying CTLs that are primed to recognize specific foreign proteins—a tool that eager researchers are now using to dissect immune responses in previously unattainable detail (Science, 4 October 1996, p. 94). “Andrew has made tremendous contributions to the field,” says immunologist Bruce Walker of the Partners AIDS Research Center in Charlestown, Massachusetts. Immunologist Douglas Nixon at the Aaron Diamond AIDS Research Center in New York City, who did his graduate work with McMichael, adds that “Andrew was instrumental in educating immunologists on the role that CTLs play in viral infections and other diseases.” In the early 1990s, however, McMichael, together with Oxford University immunologists Rodney Phillips and Sarah Rowland-Jones and Oxford biochemist-mathematician Martin Nowak, set the AIDS community buzzing when they proposed that HIV avoids being destroyed by continually mutating until it is no longer recognized by CTLs. HIV has a very high mutation rate, largely because its replication mechanism, like that of many other viruses whose genomes are made of RNA, is inefficient. This mutability, McMichael and Nowak argued, eventually allows the virus to evolve through Darwinian natural selection and escape from even the most powerful CTL onslaught. Although the concept of “immune escape” had been proposed before to explain how HIV avoids being neutralized by antibodies, assigning such a central role to CTLs was controversial, largely because it was not clear how important these cells were in combating the virus. Nevertheless, when McMichael's hypothesis was first put forward, some of his colleagues quickly latched onto it as an attractive explanation for why most HIV infections appear to be under control for many years before they eventually destroy the patient's immune system. Infected individuals produce varying levels of CTLs directed against the virus, and a growing number of researchers now believe that these cells may be key to keeping HIV at bay—and that boosting their numbers may be central to the development of a successful vaccine. But lingering doubts about the role of CTLs have made the model's key prediction—that the virus is evolving specifically in response to the natural selection “pressure” from these attacking cells—unconvincing to some researchers. The recognition of invading microbes by CTLs is a complex business. An infected cell must first chemically chop up the microbe's proteins into small fragments, or peptides. These peptides are transported to the cell surface, where they become bound to specialized molecules called human leukocyte antigens (HLAs). Special receptors on the CTL recognize the HLA-peptide complex, and the CTL then kills the infected cell by unloading a cocktail of cytotoxic chemicals into it. But even a small mutation can change the peptide's structure enough so that it will no longer bind to HLA—or, alternatively, so that the peptide is no longer transported to the cell surface—and the infected cell then becomes “invisible” to the CTL. Most researchers agree that HIV's escape from CTL surveillance could play a role in the immune system's collapse. But just how great a role is a matter of debate, and the experimental evidence brought to bear on the issue has been contradictory. “Immune escape probably has some role in a complex process that involves a large number of variables,” says Walker. “[But] a question remains as to the magnitude of its contribution to overall progression” of the disease. A dimmer view is taken by Simon Wain-Hobson, an AIDS researcher at the Pasteur Institute in Paris. Wain-Hobson, together with other European colleagues, looked for HIV “escape mutants,” but found no evidence that mutations were occurring at a faster rate in the protein segments specifically recognized by anti-HIV CTLs. Some other researchers who have looked for evidence that immune escape actually occurs have also drawn a blank. Wain-Hobson argues that the model is an “unsatisfactory” explanation for why the immune system loses control of HIV, because other fast-mutating RNA viruses, such as those that cause measles and yellow fever, can be controlled by the immune system either unaided or after vaccination. “If RNA viruses were [inherently] able to escape, we wouldn't be here talking about it,” he says. But over the past several years, McMichael and other colleagues at Oxford, including Phillips and Rowland-Jones, have continued to accumulate evidence that they believe supports the model. The group has identified a number of HIV-infected patients who had strong CTL responses against certain viral peptide sequences early in their infections. When these peptides were later altered by mutations, the patients were left with apparently weaker CTL responses directed against other sequences—a key prediction of the model. Similar evidence that viruses can escape from CTLs was reported last year in Nature Medicine by Persephone Borrow at The Scripps Research Institute in La Jolla, California, George Shaw at the University of Alabama, Birmingham, and other colleagues. Given this evidence, McMichael says, “it is hard to believe that [immune escape] doesn't have an effect” on the course of the disease. “This virus makes some of the strongest CTL responses we've ever seen; … you can't have strong CTLs without escape mutants. Invariably, there must be selection.” Now, a powerful new technique McMichael helped develop may finally resolve the debate. McMichael and other Oxford co-workers, together with Mark Davis, John Altman, and their colleagues at Stanford, have developed a highly sensitive assay to identify and quantify CTLs that recognize specific microbial peptides. The method, called tetramer staining, uses genetically engineered complexes consisting of four HLA-peptide subunits—each of which contains the specific peptide of interest—to detect CTLs taken directly from patients. The complexes bind so tightly to the CTLs' receptors that even small numbers of cells are readily detectable. Using this technique, a team led by McMichael, including Oxford researchers Graham Ogg, Rowland-Jones, and Nowak, along with Nixon and David Ho at Aaron Diamond, published in Science earlier this year the first clear evidence that higher levels of HIV-specific CTLs are correlated with lower concentrations of HIV in the blood of infected patients (Science, 27 March, p. 2103)—a finding consistent with the view that CTLs play a major role in controlling the virus. “This technology has had a tremendous impact on our understanding of the relationship between CTLs and [viral burden],” says Walker. As for how important immune escape will turn out to be in the development of AIDS, Walker says, “I am sure Andrew will be the one to sort this out.” 17. FRANCE # Duo Brings Hope of Immune Restoration 1. Michael Balter Paris—When Brigitte Autran and Christine Katlama attended medical school in Paris together in the 1970s, they had much in common: Both were specializing in infectious diseases, a field long dominated by men; both were avid and competitive skiers; and a few years later, as newly graduated doctors, they were among the first French physicians to take care of AIDS patients. Yet their paths diverged soon afterward. While Katlama continued caring for patients, Autran left clinical medicine for a career in basic immunological research. But in the early 1990s, the pair teamed up again, to lead one of France's most successful AIDS research collaborations. Over the past year, Autran and Katlama—who now both work at the Pitié-Salpêtrière Hospital in Paris—have published a series of encouraging reports showing that the battered immune systems of HIV-infected patients may recover, at least partially, if powerful combinations of drugs are used to reduce their viral burdens. Although AIDS researchers are still debating how much recovery actually takes place, most agree that the French team's findings have helped open the door to this important possibility. “[This] research has had a significant impact on the understanding of the immunopathogenesis of HIV disease,” says Mario Roederer, an immunologist at Stanford University. Immunologist Quentin Sattentau, of the Center for Immunology in Marseilles, says that Autran and Katlama have gained a place among France's leading AIDS researchers because “they are tough when they need to be and are not worried about stating their opinions in a forceful way.” The two began laying the groundwork for their recent discoveries more than a decade ago, while they were still working independently. After leaving the clinic for the laboratory, Autran began working with Pitié-Salpêtrière immunologist Patrice Debré to elucidate the role of the immune cells known as T lymphocytes in fighting HIV infection. In 1987, Autran and Debré, along with other French co-workers, published a landmark paper in Nature demonstrating that HIV-infected patients produce large numbers of killer cells, called cytotoxic T lymphocytes (CTLs), directed specifically against the virus. This key finding, simultaneously reported by Bruce Walker and his colleagues at Massachusetts General Hospital in Boston, effectively countered the views of some researchers at the time that CTLs were of little importance in the immune system's response to HIV. Meanwhile, Katlama was making her own mark on HIV research. As one of the small number of French physicians willing to devote themselves to studying the disease in the 1980s, she quickly developed a reputation as an expert on the opportunistic diseases, such as cytomegalovirus (CMV) infection, that ravage AIDS patients. And in 1985, Katlama's insistence that a patient from the Cape Verde islands off the west coast of Africa had AIDS—a diagnosis many of her colleagues doubted—led to the discovery of HIV-2, a West African variant of the virus that usually causes a milder form of the disease. Later, Katlama would emerge as an international leader in clinical trials of anti-HIV therapies. For example, her team was the first to demonstrate the efficacy of the anti-HIV drug 3TC and among the first to show that combining drugs could lead to better control of the virus. Thus when Autran and Katlama reunited several years ago, their experiences in basic and clinical research meant they were well placed to study how the immune systems of HIV-infected patients were responding to combination therapy. Their biggest breakthrough came in 1997, when, in collaboration with other French colleagues including Debré and immunologist Jacques Leibowitch at Raymond Poincaré Hospital outside Paris, the pair reported in Science that patients in advanced stages of HIV infection who were treated with combination therapy could recover some of their ability to mount immune responses against CMV and the tuberculosis bacterium—two of the most important opportunistic infections afflicting AIDS patients (Science, 4 July 1997, p. 112). In addition, after a year or so of therapy, these patients apparently begin regenerating so-called “naive” T lymphocytes, immune cells that have not yet been exposed to foreign antigens—a key criterion for immune system reconstitution. And more recently, Autran and Katlama, along with Guy Gorochov of Pitié-Salpêtrière, have demonstrated that antiviral treatments can help the crippled immune systems of HIV-infected patients, which are able to respond to fewer and fewer invaders as the disease progresses, to partially recover their capacity to respond to a wider range of invading organisms—another crucial indicator of immune system health. Autran cautions that she and Katlama have yet to demonstrate that the immune system of HIV-infected patients can be completely restored to normal, although she adds that there may be “no major limitations” to this restoration if the virus can be adequately controlled over the long term. But many researchers credit Autran and Katlama with providing new hope that this welcome possibility might become a reality. The pair's work, says Roederer, “demonstrates that there may be a slow, sustained reconstitution of the immune system … [which is] the ultimate goal of AIDS therapy.” 18. TOP LABS # A Cluster of Europe's AIDS Research Stars 1. Michael Balter European scientists have from the very beginning made major contributions to AIDS research. In addition to those whose work is described in more detail in this report, this survey highlights scientists who stand out in the opinions of their peers. It is based on several months of informal polls and discussions with AIDS researchers and is by no means exhaustive—just a sampling of the talent to be found on the continent. Marc Parmentier at the Free University of Brussels helped spark a research revolution in 1996 when he cloned a molecule called CCR5, a cell surface receptor for immune signaling molecules called chemokines, which shortly afterward was identified as a key “coreceptor” used by HIV to enter its target cells. The Pasteur Institute in Paris harbors a number of AIDS stars. Françoise Barré-Sinoussi, who first isolated HIV in 1983, has spent recent years accumulating mountains of data on African strains of the virus. Simon Wain-Hobson, who made early contributions to the molecular biology of HIV, has recently been dissecting how the virus replicates in lymph nodes, and Jean-Louis Virelizier is working on ways to block HIV binding to CXCR4, another key coreceptor for HIV. Also noteworthy are Pasteurien Olivier Schwartz's groundbreaking studies into how the HIV protein Nef down-regulates immune responses to the virus. Nearby in Paris, Marc Alizon of the Cochin Institute is studying how HIV enters cells, while down in the south of France, Quentin Sattentau at the Center for Immunology in Marseilles is recognized as a leading expert on antibody neutralization of HIV. African green monkeys do not get sick when exposed to SIV, the simian version of HIV, and Reinhard Kurth's team at the Paul Ehrlich Institute in Langen has been studying why. Their work led them to IL-16, a cytokine that has anti-HIV activity in monkeys and humans and may have therapeutic potential. Jan van Lunzen and Hans-Jürgen Stellbrink at the Eppendorf University Hospital in Hamburg have carried out laborious studies on the influence of combination therapies on viral burden and T cell numbers in blood and lymph nodes, while Andreas Meyerhans at the University of the Saarland in Homburg has helped uncover the enormous genetic variability of HIV and the dynamics of viral evolution in lymph nodes. This nation's AIDS research received a big boost a few years ago when a flock of Italian researchers returned after doing postdocs at the National Institutes of Health (NIH) in the United States. Some set up shop at the San Raffaele Scientific Institute in Milan, including Guido Poli and Elisa Vicenzi, who continue to work on the role of cytokines in the immune system disruption caused by HIV. Also at San Raffaele is Paolo Lusso, who, with Robert Gallo's former group at the U.S. National Cancer Institute, identified chemokines that block HIV infection of cells—which led to the discovery that the chemokine receptors double as ports of entry for the virus. Other former U.S. postdocs include Mario Clerici at the University of Milan, a specialist on the immune system's response to HIV, and Barbara Ensoli, whose team at the ISS in Rome—Italy's national health institute—works on Kaposi's sarcoma, a skin cancer that often afflicts AIDS patients. Dutch AIDS research is dominated by a star-studded collaboration in Amsterdam (see p. 1859), but researchers outside this orbit have also made noteworthy contributions. Jonathan Heeney at the Biomedical Primate Research Centre in Rijswijk is one of Europe's leading vaccine researchers, and Rob de Boer, at the University of Utrecht, has developed mathematical models for how HIV infection progresses. This country's leading AIDS researcher is a Swede: Birgitta Asjö at the University of Bergen. In the 1980s, Asjö carried out groundbreaking studies on HIV variability at the Karolinska Institute in Stockholm and has recently turned her attention to HIV infection of the tonsils, work that has helped researchers understand the dynamics of immune system reconstitution after treatment with combination therapies. Sweden has long been a major power in AIDS research, thanks to early pioneers such as Eva Maria Fenyö, Peter Biberfeld, and Gunnel Biberfeld, all at the Karolinska Institute. Fenyö's studies of how HIV evolves during the progression to AIDS helped establish this critical line of research, and she is now studying how viruses of differing genetic makeup differ in their use of coreceptors. The Biberfelds, who were among the first to study the pathology of tissues infected by HIV, have in recent years helped develop primate models for viral infection and are leaders in European vaccine efforts. Another Karolinska star, Britta Wahren, is trying to create a vaccine using DNA engineered to express HIV proteins. Like Italy, this alpine nation has benefited from the recruitment of European researchers who have worked in the United States, such as Giuseppe Pantaleo, now at the Vaudois Hospital Center in Lausanne after a long stint at NIH, who has accumulated key data on how the immune system keeps HIV at bay during the early period of infection. In Geneva, Luc Perrin at the University Hospital has weighed in with detailed studies of how combination therapies influence viral load in the blood of HIV patients, while Bernhard Moser and Marco Baggiolini at the University of Bern have teamed up with Jean-Louis Virelizier's group in Paris and others to develop chemokine derivatives that can block HIV attachment to its coreceptors. Since the AIDS epidemic began, Paul Clapham at the Institute of Cancer Research (ICR) in London has been one of Europe's most versatile AIDS researchers, with work ranging from his 1984 discovery with ICR's Robin Weiss that the CD4 protein was HIV's primary receptor to more recent research on drugs that block HIV binding to chemokine receptors. Studies of HIV's genetic variability by Andrew Leigh Brown and Peter Simmonds at the University of Edinburgh have helped keep the U.K. a player in this very competitive field. Finally, researchers tell Science that Angela McLean of the Institute for Animal Health in Compton—who has garnered an international reputation for developing equations that model HIV's interaction with the immune system—has saved a number of biologically trained AIDS researchers from making mathematical fools of themselves in print. 19. UNITED NATIONS # Global Program Struggles to Stem the Flood of New Cases 1. Michael Balter Geneva—Peter Piot has a seemingly impossible job. As executive director of the United Nations' special program on AIDS (UNAIDS), he is in charge of the international community's global response to the epidemic. With a staff of 130 and a budget of just$60 million a year, Piot is seeking to turn the tide against a disease that has killed more than 11 million people over the past 2 decades and is relentlessly extending its reach. Not only does he have to contend with the labyrinthine political, social, and financial vagaries of the U.N. system, but this balancing act has been made doubly difficult by the fact that his appointment coincided with a major organizational shake-up in the U.N.'s AIDS activities (Science, 25 November 1994, p. 1312). When the dust finally settled, the World Health Organization's Global Programme on AIDS (GPA), which was created in 1986, had been replaced by UNAIDS, a program jointly sponsored by six U.N. agencies, including WHO.

The new program, and Piot's appointment to run it, signaled a new era in the global battle against AIDS. Many believe the effort had been floundering since the GPA's first director—Jonathan Mann, a charismatic and outspoken epidemiologist—resigned in 1990, protesting what he called a lack of commitment to fighting the disease on the part of WHO's former director-general, Hiroshi Nakajima. Mann's replacement, Michael Merson, a public health expert who had spent much of his career at WHO, was widely criticized for unimaginative leadership. Dissatisfaction with GPA's overall performance finally sparked the chiefs of AIDS programs in other U.N. agencies to insist on an overhaul. To many observers, Piot—a soft-spoken Belgian microbiologist who cut his teeth fighting Ebola fever and other emerging diseases in Africa—brings the right blend of pragmatism and moral indignation to the job.

“In the early days there was great hope and desire to put programs into place to slow the epidemic,” says Joseph McCormick, head of the epidemiology and biostatistics unit at the Pasteur Institute in Paris. “The job of the director was to convert the skeptical and the uninformed, which needed charismatic leadership. Today the job is different, and in some ways more difficult. … It is one of convincing [political leaders] to continue to provide resources to a program for which the evidence of progress is limited.”

To be sure, UNAIDS and its predecessor have not turned the epidemic around. In fact, UNAIDS's epidemiologists estimate that 5.8 million people were newly infected by HIV, the virus that causes AIDS, in 1997 alone, and the total number of HIV-infected people is now put at more than 30 million—a much higher figure than previously thought (see sidebar). In the face of this onslaught, Piot has had to be content with more modest victories, flanking operations against an enemy whose strongest allies are poverty and the continuing complacency of many political leaders. “The biggest disappointment is the lack of political commitment in many countries, both rich and poor,” Piot told Science. “Things are happening under people's eyes, and they don't see it.” Moreover, Piot is working with reduced resources. The $60-million-per-year core budget is about 15% less than GPA had in its last years, and the staff has been trimmed considerably: GPA's roughly 275 professional employees were cut to less than half that number when UNAIDS was created, and the majority of those who remain are posted in developing countries rather than in Geneva. These economy measures reflect the desire of the sponsoring U.N. agencies to spend less supporting the day-to-day operating costs of national AIDS programs. “They made the decision that we would not be a funding agency, distributing money,” Piot says. Instead, the program aims to coordinate the work of the six U.N. agencies in each country by setting up AIDS “theme groups,” which bring their diverse activities under a single strategic plan. The theme groups provide technical aid to health authorities and nongovernmental organizations involved in battling the AIDS epidemic, as well as serve as advocates for AIDS prevention and education in each country. A cornerstone of this effort is a series of manuals published by UNAIDS, called the Best Practice Collection, which provide detailed technical advice on subjects ranging from blood safety to HIV testing to the use of the female condom. UNAIDS is also helping local authorities develop the necessary health infrastructure for expanded use of antiviral therapies, especially the treatment of HIV-positive pregnant women with the antiviral drug AZT, now that clinical trials have demonstrated that even short courses of this drug can sharply cut transmission of HIV to their infants. “UNAIDS is a very different animal than GPA was,” says Thierry Mertens, chief of WHO's AIDS and sexually transmitted diseases unit. And the reduced budget, Piot says, is not entirely a bad thing. Indeed, when he took over he discovered that more than half the money GPA had allotted for Africa had been returned unspent. “The need was there, but the money couldn't be absorbed. The AIDS programs were not well organized or managed.” Indeed, getting the six U.N. agencies—whose priorities range from UNICEF's concern for the welfare of children to the World Bank's preoccupation with development issues—to work in concert in each country is one of the trickiest parts of Piot's job. “This is like walking six cats on a leash,” says Mann, now dean of Allegheny University School of Public Health in Philadelphia. “Peter deserves a lot of credit for entering into the complex U.N. institutional environment and accomplishing what he has during the past few years.” Mann also praises Piot for continuing to stress, in his speeches and press conferences, that socially and politically vulnerable groups are most at risk from HIV infection, a point that Mann hammered home repeatedly as GPA director. “Peter has really strengthened the connection between AIDS and human rights,” Mann says. But Piot, who agrees that there is an “ethical and moral imperative to deal with AIDS,” has focused much of his efforts on making UNAIDS a “broker” to get other groups and organizations—especially private industry—to do more about the disease. “We are 17 years into this epidemic,” Piot says. “It's not enough to have morality on our side. There is also an economic reality. The price of not acting is going to cost us millions in human lives but also billions in dollars. We have to appeal to people's enlightened self-interest.” Thus while Piot travels the world trying to raise the alarm against the epidemic, his most concrete achievements have often been the result of behind-the-scenes negotiations, particularly with industry. One of his proudest accomplishments, Piot says, is the deal he struck with the Female Health Company of Chicago, the world's sole manufacturer of female condoms, to lower its prices in the developing world. The company agreed to charge less than$1.00 for the condom in poor countries, compared to its price of up to $3.00 in the industrialized world. Piot has also worked with the pharmaceutical giant Glaxo Wellcome to make the anti-HIV drug AZT available at a reduced price in developing countries, particularly to prevent transmission of the virus between pregnant mothers and their children. It is these kinds of concrete actions that have won appreciation for UNAIDS in many developing countries, where the fine points of U.N. politics might not inspire the same fascination as in New York or Geneva. Rubaramira Ruranga, an HIV-positive Ugandan AIDS activist, says that “UNAIDS helped to insure that Uganda had access to [combination] anti-HIV drugs. We are one of the few countries in Africa to have them.” Although the price of the drugs for Ugandans is high—costing each patient about$1500 each month—Ruranga, who also works as an administrator at Kampala's Joint Clinical Research Center, adds that UNAIDS is currently negotiating with drug companies to greatly reduce the price of the drugs so that many more patients can receive them. And Natth Bhamarapravati, chair of the HIV vaccine subcommittee of Thailand's National AIDS Commission, says that an expert panel organized by UNAIDS to advise his country on the wisdom of allowing the U.S. biotech company VaxGen to conduct vaccine trials “has helped Thailand to attain the maturity for making [its own] decisions. … It has helped us to be free from exploitative as well as paternalistic approaches” (Science, 30 January, p. 650).

Despite these successes, Piot says he remains preoccupied about the fate of the poorest countries afflicted by the AIDS epidemic, which cannot afford AZT, let alone sophisticated combination therapies: “This is my single biggest concern. For the least developed countries, where health expenditures are $10 or$20 per capita per year, at this point I don't see a solution. For all the other problems linked to HIV, I see the road and I see which direction to go, but for this problem I don't see it.” Until he does see the way, Piot says, in the face of the enormous tragedy brought by the AIDS epidemic, he is determined to remain realistic: “We should never make promises that we cannot fulfill.”

20. UNITED NATIONS

# HIV Incidence: 'More Serious Than We Imagined'

1. Michael Balter

Paris and Geneva—Last November, Peter Piot, executive director of UNAIDS, the United Nations' special program on the AIDS epidemic, delivered some bad news to a packed press conference in Paris: Revised estimates of the spread of HIV indicated that some 16,000 people worldwide were being infected with the AIDS virus each day, nearly twice as many as previously thought. “The AIDS epidemic is not over,” Piot said. “It is more serious than we ever imagined.”

Why were previous estimates so far off? In fact, for much of the world the numbers were pretty much what had been projected. But when epidemiologists tallied up the figures for sub-Saharan Africa, they got a nasty shock. In some highly populous countries, such as South Africa and Nigeria, the rate of HIV infection was at least twice as high as expected. The new estimates indicated that one in every eight adults was infected in South Africa, while in Botswana and Zimbabwe infection rates had reached at least 25%.

“We knew that we were underestimating the epidemic,” says Bernhard Schwartländer, UNAIDS's senior epidemiologist. “But if you had talked to epidemiologists several years ago, you wouldn't have found anyone who believed that these levels could be reached.” Schwartländer and his collaborators—who include epidemiologists at the World Health Organization (WHO), the U.S. Bureau of the Census, and the Harvard School of Public Health—were turned into believers by new HIV monitoring data that had not been available the last time similar estimates were made by WHO in 1995.

View this table:

Unlike AIDS patients who have obvious symptoms, most asymptomatic HIV-infected people have no idea that they are harboring the virus. Researchers must therefore rely on infection rates from smaller groups that have been tested for HIV—for example, pregnant women attending prenatal clinics—and extrapolate to the larger population using various fudge factors that correct for differences in age and the generally lower rates in rural versus urban areas.

In 1995, this kind of data was very sketchy for many African countries. Epidemiologists had to take what information they had and plug it into a computer model that predicted the course of the epidemic in the entire African region. This model was based largely on the dynamics of the epidemic in Uganda, which has a rigorous system of HIV surveillance. Fortunately for the Ugandans, but unfortunately for epidemiologists, the epidemic in Uganda has begun to plateau in recent years, and that led to incorrect assumptions about the course the epidemic would take in other sub-Saharan countries. In South Africa, for example, HIV infection among pregnant women had been consistently low through 1992, the last year for which data were available when the 1995 figures were released. But in reality, the epidemic was just starting to skyrocket in South Africa in the early 1990s, a fact that was painfully clear when more recent data became available.

Because HIV monitoring in many African countries has improved in recent years, Schwartländer and his colleagues were able to plug this more detailed information into computer models of the epidemic in each country rather than rely on regional models. The results left them astounded. “The doubling of incidence caused us some sleepless nights,” Schwartländer says. The researchers went back over their data to see if they had made any mistakes. But the estimates held up. “The new figures are shocking, but this is what we have to believe.”