# News this Week

Science  19 Oct 2001:
Vol. 294, Issue 5542, pp. 490
1. BIOTERRORISM

# This Time It Was Real: Knowledge Of Anthrax Put to the Test

1. Martin Enserink

Was it organized terrorism or just a madman with a grudge? Where did the attacker get the bugs? And how do you protect against anthrax anyway?

These questions were begging for answers early this week after the United States experienced what appears to have been a series of attacks with anthrax. Several contamination incidents frayed the nerves of a nation already jittery from the 11 September massacres and moved biodefense to the top of the political agenda.

As Science went to press, federal officials said anthrax-laden letters or packages had been mailed to the offices of American Media, a publishing company in Boca Raton, Florida; the NBC News desk in New York City; and the office of Senate majority leader Tom Daschle (D-SD) in Washington, D.C. Robert Stevens, a photo editor at American Media, had died of what appears to be inhalation anthrax, the most severe form of the disease, and a co-worker was diagnosed with the disease. Two other people had the milder, cutaneous form. At least eight others had been exposed but showed no signs of infection. The reports also spawned a series of hoaxes and false alarms; by last weekend, almost any powdery substance found anywhere was being treated as a potential bioweapon.

The apparent assaults posed a rare test of the country's capability to deal with a real bioterror attack—albeit a modest threat compared with the medical catastrophe that spraying a fine mist of anthrax over a big city could have wrought. But the crisis also trained a spotlight on the disease itself and the considerable investment in studying it. Thanks in part to the mounting worries about anthrax's use as a biological weapon, “there has been an explosion in knowledge,” says Martin Hugh-Jones, an anthrax expert at Louisiana State University, Baton Rouge. “It's marvelous.” Over the past 2 decades, researchers have puzzled together in detail how Bacillus anthracis makes humans sick and kills them. Even its genome of 5 million base pairs is about 95% sequenced and should be completed within a couple of months, says Timothy Read, who leads a team at The Institute for Genomic Research in Rockville, Maryland.

That molecular expertise is now being put to use on many fronts. Researchers familiar with the organism's DNA are being called on to help “fingerprint” the samples that arrived by mail in hope of identifying their origins. Others are looking at vaccines that can be administered conveniently. The old standby, the only anthrax vaccine licensed for use in the United States today, requires six shots and an annual booster. It's also in short supply, and the limited stocks are reserved primarily for military use (see p. 498). Still other researchers are developing better diagnostics to determine who is infected and who is not, as well as drugs that can block the anthrax toxin, which remains lethal even after antibiotics have killed the bacteria. All these requirements seem likely to get increased attention in the coming months.

Tough and lethal. Anthrax is a disease of livestock that occurs almost everywhere in the world. One reason it's hard to eradicate is that it forms hardy spores that can lie dormant in the soil for decades; they are found in many places in the United States. Although many researchers have worked with anthrax in animals and in the lab, precious few have ever seen anthrax in people. Even fewer have seen the pulmonary infection, caused by the same anthrax strains that cause cutaneous anthrax, provided that the spores are dispersed in minuscule particles that can descend deep into the lungs. Only 18 cases of inhalation anthrax are known to have occurred in the United States in the entire 20th century.

When the first case of the 21st century appeared this month, authorities turned to biologists for some detective work. One way to help identify the perpetrators of the attacks is to study the DNA of the spores found at the three sites and compare it to that of known strains. This could reveal whether they all came from the same source and whether they are run-of-the-mill strains available from dozens of labs or are rare varieties. Several anthrax researchers say that the FBI has enlisted the help of Paul Keim of Northern Arizona University in Flagstaff, an expert in the identification of anthrax strains. Together with Hugh-Jones, Keim has built a collection of more than 1300 strains from across the globe. Keim declines to confirm or deny his participation in the investigation, but he does point out that his lab would be better equipped than any other to do the job.

Telling anthrax strains apart is not an easy task, says Keim, because the genetic differences between strains are extremely small. One reason for their similarity may be that anthrax bacteria spend much of their time as spores, which act as evolutionary time capsules. As a result, the disease may have been around since the dawn of agriculture, but the organism has been actively evolving for only a fraction of that time, limiting its genetic variability.

Keim has developed a technique to identify different strains by focusing on a number of so-called variable-number tandem repeats, rapidly evolving spots in the microbe's genome where a small stretch of DNA is repeated multiple times. The work has already paid off in another forensic study: Keim's team was the first to identify the strain used in a 1993 anthrax attack by the Aum Shinrikyo cult in Japan. As it turned out, the cult had sprayed a nonvirulent vaccine strain into the Tokyo air, says Keim—which explains why this attack, in contrast to the later release of nerve gas in a subway, was a flop. There's no official word yet on the origins of the strains found in the United States, however.

One of anthrax's most insidious qualities is that it produces a toxin aimed at thwarting the immune system that continues to do harm even after the source is eliminated. “You can kill the bug with no effort at all,” says Hugh-Jones, “but people will still die, because they're exquisitely sensitive to the toxin.” Some researchers have focused on new ways to stop this process. For instance, Harvard University's R. John Collier, who has long been fascinated by the ingenuity of anthrax's aggressive toxin, has discovered ways to disarm it.

The toxin has three components, Collier explains. One of them, called edema factor (EF), prevents cells called macrophages from gobbling up bacteria. Another, called lethal factor (LF), kills the macrophages and eventually the host, too. The third component, protective antigen or PA (so called because it can be used as a vaccine), helps shuttle the other two into macrophages. The latter process could also be the bug's Achilles' heel, says Collier. Seven PA molecules must bind to receptors on the surface of a macrophage and come together to form a doughnut-shaped complex (see figure). Then they bind EF and LF, after which the entire complex is engulfed by the cell membrane and shuttled to a so-called endosome inside the cell. Once there, the PA molecules form a special pore that pierces the endosome's membrane and lets EF and LF out to do their grisly work.

In a paper published in Science last spring (27 April, p. 695), Collier showed that a mutated PA molecule could form part of the dough-nut like normal PA but could also disrupt the membrane pore, preventing the escape of EF and LF. Indeed, he found that rats died quickly from an injection of LF with normal PA, but survived when LF and mutant PA were injected. He hopes to create a drug based on mutant PA.

There could be a bonus, Collier says. PA is the most important component of the licensed human anthrax vaccine. Because the mutant PA elicits antibodies just as well as the normal form does, it might do double duty: “You would have wrapped into one molecule a therapeutic and a potential vaccine.” This would be valuable in a major attack, he says, when thousands of people would need immediate treatment and a vaccine to prevent infection later by lingering spores.

“It's an interesting and very important approach,” says Columbia University public health expert Stephen Morse. Harvard biologist Matthew Meselson agrees that Collier's work is “marvelous,” but at the same time, he cautions against relying on high-tech solutions to bioterrorism. Developing a new drug often takes years, if not decades, says Meselson. For now, he thinks simple, generic solutions are the best—from installing highly efficient air filters in many buildings to educating the public about do's and don'ts during an outbreak.

2. PROFESSIONAL TRAVEL

# NIH Chafes at Limits On Attending Meetings

1. Constance Holden

Some 200 scientists at the National Institutes of Health (NIH) may be no-shows next month at the year's biggest neuroscience meeting, thanks to a management directive by the Bush Administration to limit government travel. NIH officials hope that a personal appeal by several institute directors will persuade their bosses at the Department of Health and Human Services (HHS) that the policy could hinder progress in biomedical research.

“There's a lot of unhappiness,” says a source, who requested anonymity. “We're all trying to figure out ways of explaining to the Administration how pernicious this is to the process of science.”

Until recently, NIH scientists who wished to attend a scientific meeting needed clearance only from their institutes. But early this year HHS Secretary Tommy Thompson decreed that those wanting to attend meetings in foreign countries would have to get permission from his office. A few months later, the rule was extended to domestic travel. At NIH, the policy applies to groups of five or larger.

Several NIH officials—none of whom was willing to have his or her name used—told Science that the rationale doesn't appear to be financial. Instead, says one source, people in Thompson's office seem to “have the impression that traveling to meetings is a junket” rather than an essential part of the job. An HHS official says the policy is “just part of being a good steward of the taxpayer's dollar.”

The policy first showed its bite in June, when HHS lopped the list of NIH participants in the annual meeting of the Research Society on Alcoholism in Montreal, Canada, from 70 to 39. The annual meeting of the American Society of Human Genetics, which ended this week in San Diego, went on without one-third of the 170 NIH people who had asked to attend. Dean Hamer of the National Cancer Institute says that one of his graduate students was permitted to go to a “premeeting” on gene linkage but was not allowed to stay on for the main affair. “It makes no sense whatsoever,” Hamer says.

Scientists are now trying feverishly to overturn an HHS ruling that would allow only 550 of 761 researchers who requested permission to attend the Society for Neuroscience's annual meeting next month in San Diego. “Word keeps coming back to me that we are facing a stone wall,” says one source. The wall appears to be the person of Ed Sontag, recently installed as assistant secretary for administration and management. Sontag, who has a Ph.D. in education and advised Thompson on educational matters when he was governor of Wisconsin, could not be reached for comment.

NIH officials thought they were making headway this summer after Thompson visited the Bethesda, Maryland, campus and heard from postdoctoral fellows upset at the prospect of being cut out of scientific meetings. “Thompson said, ‘We need to fix this,’” reports an observer.

But nothing happened. In August, Michael Gottesman, NIH deputy director for intramural research, wrote to Thompson about the value of scientific meetings not just as a source of knowledge but also as a way to monitor grantees, recruit new talent, and nurture the careers of postdoctoral fellows. A visit by a contingent of institute directors headed by Francis Collins, director of the National Human Genome Research Institute, is also in the works. Their message will be simple, says one source: Meetings are the lifeblood of science.

3. BIOMEDICAL PHILANTHROPY

# Klausner Makes Case For New Foundation

1. David Malakoff

Richard Klausner this month traded his biomedical research battleship for a speedboat, and he's looking forward to feeling the spray on his face. His transition from the $3.7 billion National Cancer Institute (NCI)—the flagship of the National Institutes of Health—to the Case Institute of Health, Science and Technology, a small outfit in Washington, D.C., has left colleagues curious. Last week, in an interview with Science, the 49-year-old biologist for the first time laid out his plans and how they will be funded. Although the details are still being finalized, Klausner says the institute hopes to inject about$100 million over the next few years in two areas. One is a life sciences informatics initiative, aimed at developing better computerized systems for organizing and analyzing data; this will be largely in-house research. The other is a “molecular monitors” program that will disburse grants to develop technologies for identifying and monitoring specific chemicals linked to particular diseases and biological processes. The two thrusts, he says, are the “logical next steps to the things I loved to do at the cancer institute: large projects oriented toward the linkage of science and technology” that connect researchers in many fields.

With the explosion of knowledge in the life sciences, Klausner isn't worried about stepping on other funders' toes. The institute's areas of interest fall within “crowded fields,” he admits, but they offer “some novel niches.”

The new institute is funded by America Online founder Steve Case and his wife, Jean Case (Science, 14 September, p. 1967). Rather than drawing on the interest from an endowment, the institute is likely to operate on a pay-as-you-go basis, with the Cases providing fresh funds each year. The budget “would be driven by the science,” Klausner says. “I like that [approach], rather than having to worry about investing the endowment.”

The institute will soon occupy five buildings in Washington's central Dupont Circle neighborhood, with room for administrators, visiting scientists, and a significant in-house informatics program. Klausner's first top hire is MaryAnn Guerra, his former deputy director for management at NCI, who was tapped this week to be a vice president.

As for the reason he left NCI, Klausner said he relished the opportunity to lead a new organization that could change direction quickly. He also noted that instead of having scores of political bosses, he now answers to “a board of two people”—the Cases. The couple, he says, have been “very supportive” of letting him chart the institute's course.

# Swiss Source Shows Small Is Powerful

1. Giselle Weiss*
1. Giselle Weiss is a writer in Allschwil, Switzerland.

VILLIGEN, SWITZERLAND—This week Swiss researchers will proudly unveil the Swiss Light Source (SLS), a modestly sized synchrotron that punches well above its weight. Based here at the Paul Scherrer Institute, the SLS produces high-energy, or “hard,” x-ray beams comparable to much bigger sources such as the European Synchrotron Radiation Facility (ESRF) and the U.S. Advanced Photon Source. Such beams are essential for unraveling the structure of complex biological molecules and a host of other applications.

Local researchers are looking forward to having such a machine on their home patch. Tim Richmond and his colleagues at the Swiss Federal Institute of Technology in Zürich rely on x-rays to elucidate the mechanisms underlying gene expression. In earlier work teasing out the atomic structure of the nucleosome—the DNA packaging apparatus—Richmond's group had to travel to Grenoble, France, to use the ESRF. Now, they have a state-of-the-art facility just down the road. “It's the difference between being able to conceive of doing a large project and not being able to,” Richmond says, “and doing it on a reasonable time scale.”

Synchrotron light was discovered in the 1940s in early particle accelerators. When physicists bent a beam of fast-moving particles, they found that some of the energy was shed as light. The laserlike light soon proved useful in other branches of physics as well as materials science and, more recently, biology.

In the SLS, a beam of electrons is boosted to close to the speed of light in a circular accelerator and then transferred to an outer storage ring, where magnets keep the particles circulating at constant energy for between 10 and 30 hours. As the electrons round each bend, they emit light of varying wavelengths that is tapped off along beamlines to experimental stations.

As researchers get more skilled at using synchrotron radiation, they are demanding higher energies and more intensity (more photons per second). This usually means bigger machines. For example, the ESRF, which came online in the early 1990s, is 844 meters in circumference and produces an electron beam with an energy of 6 giga-electron volts (GeV). The ESRF uses specialized magnetic devices known as undulators and wigglers to manipulate the electron beam and customize the x-rays to researchers' needs.

For the SLS, project leader Albin Wrulich says, designers used a few “tricks” to get similar performance out of the new machine, pushing current undulator technology to the limit to increase x-ray intensity and the range of wavelengths available. As a result, although the SLS is just 288 meters around and has a 2.4-GeV electron beam, it produces x-rays of almost the same intensity as the ESRF's and, at $89 million, cost less than a third as much. After its 19 October inauguration, engineers will continue to fine-tune the SLS before it is opened to the research community in January 2002. For Richmond, who hopes to show people what DNA looks like in higher cells through x-ray crystallography and to relate that structure to its function, the SLS is the perfect tool. All he needs now, he says, is a crystal. 5. WEATHER FORECASTING # Getting a Handle on The North's 'El Niño' 1. Richard A. Kerr The lowermost layer of Earth's atmosphere—the troposphere, the place where all the people live—is a forecaster's nightmare. Weather patterns are capable of jumping without respite or warning from one mode of operation to another. This volatility is especially dramatic in winter. Frigid winds and powerful storms might first range far south of their usual haunts then retreat farther northward than usual, seemingly unpredictably, producing a midwinter respite. Only El Niño had seemed able to lock the weather into one regime or another long enough for forecasters to anticipate prolonged periods of extreme winter weather weeks or months ahead. Now forecasters have the prospect of another, unlikely steadying influence on the weather: the wispy stratosphere overlying the troposphere. On page 581 of this issue of Science, meteorologists Mark P. Baldwin and Timothy J. Dunkerton of Northwest Research Associates in Bellevue, Washington, report that despite its reputation as a lightweight, the stratosphere at times reaches down through the troposphere to push Northern Hemisphere weather toward one extreme regime or the other for a couple of months at a time. That could give forecasters an edge in long-range predictions at middle and high latitudes comparable to that provided by El Niño at some latitudes. And it clearly shows that “there's no brick wall up there” sealing off tropospheric weather from the rest of the atmosphere, says meteorologist Marvin Geller of the State University of New York, Stony Brook. “To understand weather or climate, it's good to look at how the whole atmosphere is behaving.” Atmospheric scientists had long identified higher latitude weather oscillations in the stratosphere and troposphere, but until now they hadn't established a downward connection between the two. In the stratosphere, a polar vortex of high-speed winds that swirls around the North Pole in a ring 10 kilometers or more above Canada, Scandinavia, and Siberia waxes and wanes, sometimes abruptly. At the surface, the North Atlantic Oscillation (NAO)—a wobbly seesaw of varying atmospheric pressure that spans from Iceland to Lisbon—swings from one extreme to the other, redirecting winds around the North Atlantic and switching regional weather between cold and stormy and mild and fair (Science, 7 February 1997, p. 754). In 1998, meteorologists David Thompson of Colorado State University in Fort Collins and John M. Wallace of the University of Washington, Seattle, expanded the NAO concept to encompass the entire higher latitudes and called it the Arctic Oscillation (AO). The tropospheric AO works much as the polar vortex does in the stratosphere. The AO involves the prevailing westerly winds that oscillate in strength and position to shift weather patterns around the hemisphere (Science, 9 April 1999, p. 241). When the shifts persist long enough, climate changes. Thompson and Wallace's work prompted Baldwin and Dunkerton, who are stratosphere specialists, to look higher in the atmosphere for AO connections. To their surprise, they found downward links as well as upward ones. A switch from a strong stratospheric vortex to a weak one, say, would move down through the stratosphere, entering the troposphere and reaching the surface as a weakening and diversion of the AO's westerly winds. Because it took a few weeks for a switch to get from the vortex to the AO, predicting a switch a week or two ahead looked possible. Baldwin and Dunkerton have now taken a more detailed look at 42 years of vortex and AO wintertime behavior and found that the connection can be a persistent one. Once a major switch reaches the lower stratosphere, the vortex remains unusually weak or strong for an average of 60 days, which should let forecasters predict extremes in the underlying AO and the accompanying likelihood of weather extremes out as far as a month or two. Forecasters might, for example, warn that cold air outbreaks from the Arctic into midlatitudes would be three to four times more likely across Europe, Asia, and North America. In separate, as-yet-unpublished analyses, Thompson, Baldwin, and Wallace find that major vortex and AO shifts affect surface temperatures about as much as El Niño does. In central Europe and most of North America, surface temperatures average 0.5° to 2°C cooler in the 60 days following the onset of an extremely weak vortex than in the same period following the onset of a strong vortex. The difference is 1.5° to 4°C for the high Eurasian Arctic. That compares with temperature differences between El Niño and its opposite, La Niña, of 1° to 3.5°C in higher latitudes. “The El Niño analogy is a good one,” says Baldwin. “The magnitude of these [switches] could be very useful.” Researchers are generally impressed. The Baldwin and Dunkerton “analysis is very careful and very complete,” says stratosphere meteorologist Karin Labitzke of the Free University Berlin. Predicting weather based on the work will be harder, as Baldwin and Dunkerton point out, because switches in the stratosphere and the AO sometimes occur independently, and no one understands the mechanics of the stratosphere-troposphere linkage when it does happen. Forecasters' computer models “must be able to predict where and when the effects of this interaction [between stratosphere and troposphere] will be manifested,” says Edward O'Lenic, a long-range forecaster at the National Weather Service's Climate Prediction Center in Camp Springs, Maryland. “This is a tall order and a challenge for modelers, but the payoff could be great.” Modelers are already trying to sort out how the stratosphere can influence the weather. The stratosphere might gain leverage on the troposphere through great globe-girdling atmospheric waves that rise into the stratosphere during winter. How the stratosphere and troposphere communicate will be of interest not only to long-range forecasters, but to climatologists as well. The same linkage may well be operating when volcanic debris, an inconstant sun, ozone depletion, or greenhouse gases alter stratospheric climate. Perhaps more than one forecasting nightmare could be eased by understanding stratospheric harbingers. 6. VOLCANOLOGY # Vesuvius: A Threat Subsiding? 1. Alexander Hellemans* 1. Alexander Hellemans is a writer in Naples, Italy. NAPLES, ITALY—People living in the shadow of Vesuvius, the volcano that so famously buried the Roman town of Pompeii, may be able to sleep a bit easier. New satellite data, some experts say, suggest that the small earthquakes that shake the region almost daily are not harbingers of an imminent eruption. Rather, they occur because the central part of the volcano's crater is sinking at a rate of several millimeters per year. About 1 million Neapolitans might have to be evacuated if Vesuvius awakes from its 57-year-long slumber. Scientists and civil defense experts are bitterly divided over the adequacy of evacuation plans. There is no way of knowing when Vesuvius might erupt again, but rising magma beneath active volcanoes can produce tremors before an eruption. To aid the debate, Riccardo Lanari and his colleagues at the Research Institute for Electromagnetism and Electronic Components in Naples used radar interferometry to look for tiny ground deformations around the volcano. The European Space Agency's ERS-1 and ERS-2 satellites took radar images of the Vesuvius region repeatedly between 1992 and 2000. The images recorded microwaves that were emitted by the satellite and bounced back from the ground below. By comparing images taken at different times, researchers can spot small deformations in the level of the terrain. For the current study, Lanari's team used newly developed software to detect minuscule deformations over shorter periods than previously possible. The team identified two key areas of subsidence, both of which are slumping by 3 to 8 millimeters per year. One is a ring circling the base of the volcano, whereas the other is within the central crater. “This is quite a spectacular image,” says Steve Sparks of the University of Bristol, U.K. Local researchers believe that the ring-shaped deformation is caused by the weight of the volcano on the underlying rock, the so-called carbonatic basement. The radar observations confirm earlier suggestions of this slumping using seismic tomography—bouncing sound waves off underlying rock. “We were able to see the spectacular effects of the volcano loading,” says Giuseppe De Natale of the Vesuvius Observatory. More intriguing is the subsidence in the central crater. Researchers had thought that the frequent minor tremors in the area—which seemed to come from directly below the crater—were due to molten magma welling up. But according to De Natale, the central subsidence confirms that a giant magma plug—a “high-rigidity anomaly” about 5 kilometers long—is sinking. The tremors are caused by the plug grinding against surrounding rock. Sparks thinks this makes sense. “It is sort of consistent with what people have been deducing from model experiments and theory,” he says. Although the technique is impressive, says Paul Lundgren, an earthquake and volcano specialist at NASA's Jet Propulsion Laboratory in Pasadena, California, more work is needed. The result “is not high enough above the noise that you get a really good crystal-clear picture of what is going on,” he says. De Natale adds that radar and other geodesic methods may be useful for forecasting eruptions: Significant uplift could indicate that Vesuvius is waking from its sleep. 7. ASTRONOMY # Pulsars Solve Mystery Of Missing Gas 1. Govert Schilling* 1. Govert Schilling is an astronomy writer in Utrecht, the Netherlands. By rights, clusters of stars should be swimming in gas. Stellar winds just like the solar wind blowing from our sun should fill the space between their stars. Yet decades of searches have failed to turn up interstellar gas in globular star clusters, and theorists have bent over backward to explain its absence. Now they can unbend. Meticulous observations using the 64-meter Parkes radio telescope in Australia have revealed that the globular cluster 47 Tucanae harbors about 100 times as much gas as its immediate surroundings. “It's a very nice discovery,” says astrophysicist Frank Verbunt of Utrecht University in the Netherlands. Globular clusters are spherical clumps of hundreds of thousands of old, low-mass stars, held together by the mutual pull of their gravity. To understand how such clusters evolve, astronomers want to know how much of the stellar wind escapes the cluster and how much remains trapped by gravity. “If the stellar wind velocity is larger than some 50 kilometers per second, the gas will leave the cluster altogether,” says Verbunt. Numerous searches for radiation coming from trapped gas, however, revealed nary a whiff, and some astronomers proposed that starlight had blown it all away. “It has been suggested that the energetic radiation of the large number of pulsars in 47 Tucanae would have rendered the cluster devoid of any gas,” says Michael Kramer of the Jodrell Bank Observatory at the University of Manchester, U.K. Kramer and an international team of astronomers led by Paulo Freire, also of Jodrell Bank, detected the gas by carefully observing 15-millisecond pulsars—very compact, rapidly spinning stars that emit bursts of radio waves with clockwork precision. As the pulsars orbit the cluster center, Kramer explains, gravitational acceleration minutely alters the pulse frequency. By monitoring such changes over a couple of years, the team determined whether each pulsar lay on the near side of the cluster center, where its acceleration points away from us, or on the far side, where it points toward us. “We didn't expect the data to be precise enough [to observe this effect],” says Kramer, “but it turned out we could track it.” Next, the astronomers observed the radio pulses at a variety of wavelengths and found that the radiation from the farside pulsars was affected by more interstellar gas than was the radiation from the nearside ones. From these differences, they calculated the density of the intracluster gas. “People have been looking for this gas for over 60 years,” says Kramer. “Now we've found it.” The team reported its discovery in the 10 October issue of Astrophysical Journal Letters. There's not much gas there, however: about one-tenth of a solar mass, spread out over the central parts of the cluster. The next step for astronomers is to figure out why it is so thin. Did most of it escape as high-velocity solar wind, did pulsars blow it away, or was it stripped out when the cluster made one of its periodic passes through the dense central plane of the Milky Way? Verbunt plans to look for answers in another globular cluster using the Westerbork Synthesis Radio Telescope in the Netherlands. 8. BIOTERRORISM # Vaccines for Biodefense: A System in Distress 1. Jon Cohen, 2. Eliot Marshall Terrorist attacks highlight the need for exotic new shots; development has lagged, thanks to a lack of commercial interest and management snafus Two months before the 11 September terrorist attacks, the U.S. Department of Defense (DOD) sent Congress a report that attracted little attention at the time but has suddenly been catapulted into prominence. An unsparing and sometimes scathing critique by an independent panel of experts, the report bluntly concludes that the military's system for developing vaccines to protect troops from anthrax, smallpox, and other exotic bioweapons “is insufficient and will fail.” The document has become a hot topic in Washington, D.C.—from the White House on down. It has been seized upon by scientists who have long been arguing that the United States is ill prepared to defend against bioterrorism, and it helps answer a question that has become pressing in the wake of last week's anthrax scare (see p. 490): Why, several years after experts began issuing dire warnings about bioterrorism, are there no new vaccines against some of the most worrisome potential agents and scant supplies of existing vaccines? The report, sources tell Science, may lead to a sweeping overhaul of how the federal government develops vaccines to protect both the military and civilians from bioweapons. But the problems are not just logistical: Developing a new suite of vaccines presents difficult scientific challenges, and testing the safety and effectiveness of preparations against lethal diseases that aren't already widespread will be problematic. “In light of the events of September 11, [the issue] takes on a new sense of urgency,” says Anna Johnson-Winegar, a microbiologist who oversees and coordinates DOD's biological and chemical defense program. Vaccines form the cornerstone of DOD's extensive program to defend against biological attack (see table below). Industry has little financial incentive to develop such vaccines, so the Pentagon, until recently, has been virtually alone in funding their development. In the mid-1990s, DOD tried to move things along by creating the Joint Vaccine Acquisition Program (JVAP). It takes promising biodefense leads from military researchers and hands them to an outside contractor, which then farms out vaccine production to other contractors. Currently, JVAP has targeted eight vaccines against bioweapons. The independent panel called this arrangement cumbersome and poorly coordinated, noting that the number of organizations involved “seems unnecessary and counterproductive.” The report “was not pretty to write,” says the panel's chair, Franklin Top, who previously served as the commander of the Walter Reed Army Institute of Research (WRAIR) and now is an executive at MedImmune, a biotechnology company in Gaithersburg, Maryland. Top and the four other panelists recommended a radical gold-plated solution: Scrap JVAP and replace it with a$3.2 billion military program that would produce its own vaccines in a government-owned production plant.

Outside researchers applaud the idea to scrap JVAP. “It really is a terrible operation,” says epidemiologist D. A. Henderson, who formerly headed the World Health Organization's (WHO's) smallpox eradication program and now directs the Center for Civilian Biodefense Studies at Johns Hopkins University. Maj. Gen. Philip Russell (retired), another former head of WRAIR, uses even stronger language: “It's a disaster.”

The idea for a huge government-run vaccine production operation surfaced after the Persian Gulf War, but Congress and DOD rejected it as too expensive. 11 September changed all that. Now, it's being considered for producing vaccines to protect not just troops but civilians as well. “It's being discussed in a manner of seriousness that I haven't seen before,” says Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases (NIAID), who says he personally likes the idea. “Sometimes you've just got to take things into your own hands and go for it,” says Fauci. “There's a real and present danger of bioterrorism. The government needs to say enough is enough … we're going to make sure these things get made.”

DOD's independent panel did not discuss problems with specific products, but Science looked closely at DOD's efforts to develop vaccines against three of the most feared bioweapons: smallpox, anthrax, and botulinum. Each illustrates why, in the words of vaccine researcher Leonard Smith of the U.S. Army Medical Research Institute of Infectious Diseases (USAMRIID) in Fort Detrick, Maryland, officials at the Office of the Secretary of Defense “are pulling their hair out trying to think what they can do.”

## Smallpox

Once the scourge of humankind, smallpox has been so well controlled by vaccination that WHO declared in 1980 that it had been eradicated. The only known remaining samples of the virus are secured in scientific freezers at the U.S. Centers for Disease Control and Prevention (CDC) in Atlanta, Georgia, and at the State Research Center of Virology and Biotechnology in Novosibirsk, Siberia. Because smallpox vaccination itself carries some risk, countries also decided to end routine vaccination. But after the Soviet Union collapsed, fears that terrorists might get their hands on some of the Novosibirsk smallpox led some experts to urge the stockpiling of smallpox vaccines.

The United States has 15 million doses of aging smallpox vaccine—too few to provide adequate military and civilian protection, according to most experts. So in 1997, DOD's JVAP contracted with DynPort Vaccine Co., a British-U.S. joint venture based in Frederick, Maryland, to make 300,000 doses of a new smallpox vaccine for the military. At about the same time, President Bill Clinton became interested in stockpiling smallpox vaccine to protect the public as well.

Scientists wanted to develop something better than the old-fashioned smallpox vaccine, made by scraping vaccinia, a viral cousin of cowpox and smallpox (variola), onto the bellies of calves and harvesting the pustules that formed. Researchers at USAMRIID, the lead military lab dedicated to developing defenses against biological weapons, decided to grow vaccinia in tissue cultures for greater purity—and quickly hit a barrier.

Although Wyeth Laboratories was the last manufacturer to produce smallpox vaccine, USAMRIID researchers, for reasons that no one can now recall, turned to another company, Connaught, for vaccinia. From an array of vaccinia in the old preparation, they selected clones that were less “reactogenic” (produced fewer harsh reactions) in rabbits. But researcher Peter Jahrling, a USAMRIID virologist, acknowledges that researchers do not yet have evidence that the new vaccine will be less reactogenic in humans. Reactogenicity “is going to be a problem,” predicts Top, who thinks the vaccine may cause side effects similar to those of the old one. Although rare, the side effects can be severe, including brain swelling, aggressive eczema, and—in people who suffer from immune damage such as those infected with HIV—a dangerous pox infection.

Jahrling had practical concerns as well: It was taking too long. USAMRIID had sent the tissue-grown vaccinia to JVAP in 1995, but DynPort started working on the project only in 1998. “There was a lot of hand wringing and gnashing of teeth about how vaccine will be made,” says Jahrling. DynPort's president, Terry Irgens—a former commander of the Naval Medical Logistics Command at Fort Detrick—declined to discuss the smallpox vaccine, saying “we've been strictly forbidden” to talk about it.

The slow pace of vaccine development frustrated government officials and researchers alike, say both Henderson and Jahrling. Richard Clark, the national counterterrorism czar under Clinton, became involved. Jahrling says Clark wanted DynPort to provide vaccines for the Department of Health and Human Services (HHS) for civilian use, as well as for DOD. “Clark said, ‘Listen, damn it, we are not going to have two vaccines; we're going to have one,’” recalls Jahrling. A big sticking point, say both Henderson and Jahrling, was the cost: DOD was paying DynPort about $75 per dose of vaccine, whereas HHS wanted to pay only about$1. In the end, HHS decided to select its own manufacturer: Last September, CDC awarded a 20-year, $343 million contract to OraVax, a Cambridge, Massachusetts, biotech. The price:$1.38 per dose. Both DynPort and OraVax subcontracted with yet another outfit, BioReliance of Rockville, Maryland, for help.

When OraVax asked JVAP for the seed vaccinia, however, DOD balked, claiming liability concerns. (OraVax, which in December 2000 changed its name to Acambis, declined comment, saying CDC forbids the company from discussing its smallpox vaccine program; CDC denied Science's request to speak with the agency's leading expert about smallpox.) One person close to the negotiations says that OraVax's lawyers offered to indemnify DynPort, but JVAP still refused to provide the seed vaccinia. Subsequently, the company obtained seed vaccinia from Wyeth.

Today, DynPort and Acambis are making smallpox vaccines that are essentially the same—both with help from BioReliance. Critics suggest that DOD should simply abandon its plan to have DynPort make 300,000 doses—an amount that both Jahrling and Henderson say is inadequate to protect U.S. troops—and purchase the cheaper vaccine from Acambis.

After 11 September, Acambis, at CDC's request, said it could step up its production schedule and have vaccine ready by summer 2002. Plans are also under way to have the company make tens of millions of doses more than scheduled.

In the meantime, Fauci says NIAID scientists will soon launch studies to see how much they can dilute the 15 million doses of old Wyeth vaccine in the stockpile without reducing the ability to stimulate a robust immune response. “There's the potential to go from 15 million doses to 150 million,” Fauci says. NIAID also expects to help CDC with clinical testing needed to win approval from the Food and Drug Administration (FDA) for the new Acambis vaccine. And sources tell Science that officials are considering compelling large pharmaceutical companies to manufacture up to 300 million doses of smallpox vaccine and make them available within a few months as an unlicensed “investigational new drug.”

## Anthrax

Unlike smallpox, anthrax has become a real and present bioterror threat. But the good news is that the FDA has already licensed a vaccine, and animal tests have shown that it protects against the worst infections: those caused by inhaled spores. The bad news, however, is that only one company makes the vaccine, BioPort of Lansing, Michigan, and BioPort's production line has been down since 1998.

Before the company changed hands that year, FDA inspected its facilities and found them deficient. Production was suspended while renovations were begun, but according to testimony from a BioPort executive in the House of Representatives last year, the new owners found that the required improvements were more costly than they had anticipated. Even after DOD agreed to double the contract price of the vaccine, the company was struggling to survive.

Before it stopped production, BioPort had accumulated more than 2 million doses of vaccine. All have been spoken for. Some were quarantined on FDA's orders. About 1.8 million were used by the Pentagon during the Clinton Administration for a mass vaccination of troops, and the remaining 24,000 doses “are dedicated to the military,” says Lt. Col. John Grabenstein, deputy director of DOD's Anthrax Vaccine Immunization Program. Vaccinations are being given only to “special mission units and researchers” in high-risk labs, he says. BioPort is submitting an application this month to FDA to resume production, and an agency spokesperson says the review will be conducted soon, because “we're not handling this as a routine inspection.”

BioPort's vaccine is a complex broth of proteins filtered from a nonthreatening strain of anthrax. USAMRIID has been pursuing an alternative vaccine based on a genetically engineered version of one key antigen. The leader of this research, Col. Arthur Friedlander, says, “We're working with NIH on a fast track to get this into clinical trials.” He acknowledges that researchers still don't know just how the vaccine works or whether the recombinant version will be as effective as the old broth. But he says that studies with “the best animal models,” including nonhuman primates, indicate that results are “essentially the same” as for the old vaccine: 95% are protected, even against inhaled anthrax. DOD's Johnson-Winegar calls it “a very promising candidate,” adding that it should begin clinical trials “early next year.” CDC, meanwhile, is conducting a 1300-person trial of the licensed vaccine to learn whether the burdensome schedule can be cut from six shots to five.

## Botulinum

Smallpox and anthrax may be getting the most attention, but biodefense researchers have long been concerned about botulinum toxin, the deadliest of all potential toxic threats gram for gram. Produced by an anaerobic bacterium that can grow in ill-preserved food (Clostridium botulinum), it attacks the cholinergic nerve system, causing death by paralysis. Ingesting less than a millionth of a gram can be fatal. To fully protect against botulinum, a vaccine would have to trigger antibodies against all seven known strains, or serotypes, of the bacterium.

Botulinum toxin has been viewed as a potential bioweapon for decades; indeed, when an international arms control team swept Iraq in the mid-1990s, it found that Saddam Hussein's labs had churned out 19,000 liters of serotype A botulinum toxin and loaded some of it into warheads.

Defending against all possible botulinum weapons will be difficult, researchers say. But according to USAMRIID's Smith, public health and military researchers developed a vaccine in the 1970s that offered relatively broad protection. It uses modified versions of five toxins (called toxoids) to stimulate antibody protection against five serotypes. Omitted are two serotypes considered to be less toxic or difficult to manufacture. However, FDA has not licensed this “pentavalent” vaccine, and it remains difficult and dangerous to mass produce, requiring a dedicated manufacturing facility. Its crude mix includes formaldehyde and other components that make it highly reactogenic.

A new botulinum vaccine has been in the works now for more than a decade. Researchers at U.S. Army labs have used recombinant technology to genetically engineer yeast factories that produce four partly dismantled botulinum toxins of serotypes A, B, C, and F. Serotype E will be added soon. The resulting combination, says Smith, is a candidate to replace the old vaccine: one that could be manufactured more safely and inexpensively.

Botulinum vaccine experts are eager to begin clinical trials, but progress has been slow. The developmental project must now be carried forward by JVAP and its contractor, DynPort. “It's absolutely astounding” how cumbersome and expensive the process has become, says one observer. The recombinant serotype B vaccine was ready for testing 4 years ago, according to an expert, and “it could have gone into phase I trials then.” Although military researchers have been in consultation with FDA for 2 years planning safety and efficacy tests for all four serotypes, clinical trials are still “nowhere in sight,” a researcher says.

## Fast track

Even if JVAP can produce new biodefense vaccines, all of them face a huge obstacle: Because the diseases they are designed to block do not circulate in the population, there is no way to stage classical efficacy trials in humans. Samuel Katz, a pediatrician and professor emeritus at Duke University—who sits on the Institute of Medicine's panel on military vaccines—says this is not a trivial problem. He points to the new smallpox vaccines as an example. Although researchers have long used measurements of antibodies and skin reactions to determine whether a smallpox vaccine is immunogenic, no one knows precisely which immune responses lead to protection. “Is this new vaccine more immunogenic, less immunogenic, more efficacious, less efficacious?” asks Katz. “I don't know how you're going to answer those questions.”

FDA recognizes this dilemma and in 1999 issued a proposed rule that would allow the agency to license vaccines and drugs against bioweapons without human efficacy studies. FDA would base approval instead on “substantial evidence” from studies in two animal species but would pull such a product from the market if evidence surfaced that it did not work in humans. FDA says it does not know when it will make a final decision on the proposed rule.

Given all the uncertainty about bringing new biodefense vaccines to market and the myriad of federal agencies involved, a radical overhaul of the system seems likely. Since the terrorist attacks, says Russell, “people have begun to think about the total national interest rather than their bureaucratic territory.” DOD's Johnson-Winegar is chairing a high-level government panel that, she says, is reviewing how to invest in a “national resource” for military and civilian vaccines; a dedicated government facility is “one of the options” being considered. That brings the discussion back to where it stood a decade ago, at the end of the Gulf War.

9. BIOTERRORISM

# Blocking Smallpox: A Second Defense

1. Jon Cohen

Variola, the virus that causes smallpox, was destined for obliteration 6 years ago when a select group of researchers took a new interest in it. The team, spearheaded by Peter Jahrling and John Huggins of the U.S. Army Medical Research Institute of Infectious Diseases (USAMRIID) in Fort Detrick, Maryland, extensively studied a promising antiviral treatment for variola infection that might be useful in containing an outbreak. They also demonstrated for the first time that the virus can cause fulminant disease in a species other than humans. But their work has been controversial, in part because it raises a question about the World Health Organization (WHO) plan to destroy all known stocks of variola.

For security reasons, the experiments took place in Atlanta at the Centers for Disease Control and Prevention's (CDC's) biosafety level 4 laboratory, which has special air filters that allow researchers to handle the most lethal viruses on Earth. “It's kind of a thrill to be working with [variola],” says Jahrling. “It's an awesome responsibility.”

In 1995, Huggins and his co-workers, including James Leduc of the CDC, decided to pursue an antiviral lead from Erik De Clercq and Johan Neyts of the Rega Institute for Medical Research in Leuven, Belgium. Two years earlier, the Belgian researchers had reported that in mouse experiments, a drug known as cidofovir—made by Gilead Sciences of Foster City, California—could thwart vaccinia virus, a tame vaccine strain. In test tube experiments, Huggins and colleagues found that cidofovir worked remarkably well against 31 different strains of variola, too. They also showed that cidofovir can protect monkeys exposed to a close relative of variola, monkeypox. And they identified 10 compounds that appeared even more potent than cidofovir.

Donald Smee, a virologist at Utah State University in Logan who has collaborated with the USAMRIID researchers, notes that cidofovir has two attractive features: It works with a single dose, and it already has been licensed (although for a different use) by the U.S. Food and Drug Administration. If used in conjunction with vaccine during a smallpox outbreak, says Smee, “the damage could be certainly contained more quickly.”

Cidofovir does have a downside: It must be given intravenously. “If you had a biological attack where a lot of people had to be treated, IV [would] become a pain,” says Smee, who, along with De Clercq, notes that it might be relatively easy to make an aerosolized form of the drug. Epidemiologist D. A. Henderson, the former head of WHO's smallpox eradication program, says he's “very doubtful” that an antiviral drug would prove effective against symptomatic smallpox. Nor does he think it would be preferable to a vaccine during the window of time between infection and disease, when the vaccine could still help. Henderson says, however, that antivirals like cidofovir could play a useful role in treating problems such as infection of immunocompromised people caused by the renewal of vaccinations with the vaccinia virus (see main text).

It will be difficult to prove that cidofovir works against smallpox, however, because the disease was declared eradicated in 1980. This is where Jahrling's monkey studies could be critical. Jahrling and his colleague Lisa Hensley last year found a strain of variola that could infect and cause severe disease in cynomologous monkeys—a first. “They had vesicular lesions that looked just like smallpox,” says Jahrling. In all, 11 of these 12 monkeys died within a week to 10 days. This is much “hotter” than human variola, which takes a few weeks to cause disease, and Jahrling now is attempting to refine the model. But Henderson calls the monkey model “ridiculous,” because he is not convinced that it represents human smallpox. “There's a great desire on the part of [the Department of Defense] to justify retaining the smallpox virus,” says Henderson. “And that's what it all boils down to.” Henderson is referring to the fierce debate about whether the remaining stocks of variola should be destroyed. If a monkey variola model proves valid, it could alter the decision. Jahrling plans to present his data this December at a WHO meeting about variola destruction; unless the plan changes, variola will be snuffed out in December 2002.

10. NOBEL PRIZE IN PHYSIOLOGY OR MEDICINE

# Cycling Toward Stockholm

1. Michael Balter,
2. Gretchen Vogel

Three new laureates per prize—the maximum number Nobel rules allow—gain recognition for fundamental advances in their fields

LONDON—Paul Nurse and Tim Hunt have been good friends for nearly 20 years. Indeed, they have much in common. Both have conducted pioneering research into the intricate molecular choreography that drives cell division. Both work for Britain's Imperial Cancer Research Fund (ICRF): Nurse as its director-general and Hunt as head of ICRF's Cell Cycle Control laboratory in South Mimms, north of London. “We are very complementary,” Nurse tells a visitor who has come to see the pair in Nurse's London office, overlooking the lush green gardens of Lincoln's Inn Fields. Now, Nurse and Hunt have something else to share. With yeast geneticist Leland Hartwell, director of the Fred Hutchinson Cancer Research Center in Seattle, Washington, they have been awarded this year's Nobel Prize in physiology or medicine for identifying the key molecular steps in the cell cycle.

The winners' work led to the revelation that the cell cycle is controlled through the cooperation of two sets of proteins: the cyclins and enzymes called kinases. Their discoveries not only illuminated cell biology's most fundamental process—the ability to grow and divide—but also have had important implications for medicine. “The principal problem in cancer cells is they divide when they shouldn't,” says cell biologist Ted Weinert of the University of Arizona in Tucson. “Without these discoveries, cancer research would still be in the dark ages.”

Things were dark indeed in the late 1960s, when Hartwell began his work in yeast genetics. “People knew there was a cell cycle, but as for how to get at the genes, it didn't cross anyone's mind that it was even possible at that point,” Weinert says. Hartwell's research sprang from a project he had assigned to Brian Reid, an undergraduate student in his new lab at the University of Washington, Seattle. “I gave him mutant [yeast strains] that formed very odd shapes at high temperatures,” Hartwell recalls.

The oddly shaped cells, it turned out, were having trouble dividing. After one or two generations, for example, each would have a very large bud, which had failed to separate and become a daughter cell. “We were immediately stunned by the amount of information they gave us on cell division,” Hartwell says. They could see the physical consequences of the mutation as well as at what point in the cell cycle the mutation exerted its effect. By the early 1970s, the Hartwell lab had identified dozens of gene mutations that disrupt the cell cycle, although the nature of the proteins encoded by these “cell division cycle” (cdc) genes wasn't known.

While Hartwell was making his seminal discoveries, Nurse was completing his graduate studies in amino acid metabolism at the University of East Anglia in Norwich, U.K. But the lab's amino acid analyzer kept breaking down, and Nurse had to spend many hours babysitting the machine. This gave him plenty of time to read journal articles, including Hartwell's early papers. “I saw that this genetic approach is really powerful,” Nurse says.

After receiving his Ph.D., Nurse began working at the University of Edinburgh, U.K., searching for cell cycle genes in the fission yeast Schizosaccharomyces pombe. Like Hartwell, he soon identified a number of cdc genes, as well as so-called “wee” mutations that caused the yeast to go into mitosis early, stunting the growth of the cells. When one of these mutated genes, wee2, turned out to be a mutant form of a gene Nurse had isolated earlier called cdc2, he reasoned that cdc2 must control when mitosis begins.

Subsequent work revealed that cdc2 operates as a master control switch that determines the timing of key steps in cell division. Moving on to the University of Sussex, and later to Oxford and London, Nurse and his co-workers found that the cdc2 gene codes for a protein called a kinase, part of a family of regulatory enzymes important to many cell functions. The group also showed that cdc2 is nearly identical to Hartwell's cdc28 gene in bakers' yeast.

In the meantime, Hunt, who had received his Ph.D. from Cambridge University, began spending summers at the Marine Biological Laboratory in Woods Hole, Massachusetts. He was trying to figure out how fertilization of the sea urchin egg triggers protein synthesis, one of the first steps in the development of the sea urchin embryo. But the work seemed to be going nowhere. Then, in 1982, Hunt tried what he now calls “a completely off-the-wall” experiment “of the desperate variety.”

He decided to compare the protein synthesis patterns in fertilized eggs with those that developed parthenogenetically—that is, without fertilization. The results, he says, were a “complete revelation.” In the fertilized eggs, the levels of a protein present in high concentrations dropped drastically just when the cells divided. Then the levels rose again, only to drop at the next round of cell division. Hunt named this protein cyclin, the first of many such proteins to be discovered.

It soon emerged that Hunt and Nurse were independently looking at two facets of the same problem. Further work showed that cyclins regulate the enzymatic activity of the Cdc2 protein and other so-called cyclin dependent kinases (CDKs). In fact, Cdc2 and its cyclin regulators actually join together to form a larger molecular complex called maturation promoting factor (MPF). MPF—which had first been identified as the key initiator of cell division in frog eggs in the early 1970s by the Japanese scientist Yoshio Masui—had long resisted biochemical analysis. Now, that mystery was solved.

In 1987, when Nurse isolated the human version of Cdc2—called CDK1—it also became clear that the cell cycle is controlled by a universal mechanism that has been conserved in yeast, amphibians, mammals, and other organisms over nearly 2 billion years of evolution.

Since then CDKs, cyclins, and their associates have been at the center of research on both normal and cancerous cell growth. “If you said, ‘Let's give a prize for CDK,’ these are the three people you would give it to,” says cell biologist Kim Nasmyth of the Research Institute of Molecular Pathology in Vienna, Austria. “I think [the Nobel committee] got it absolutely right.” In fact, says cell biologist Joan Ruderman of Harvard University, who worked with Hunt on some of his early studies, the whole field is enjoying the moment in the spotlight. “For many of us, it feels like cyclin/Cdc2 has won the Nobel Prize, and we are all very happy about that!”

11. NOBEL PRIZE IN PHYSICS

# Laurels for a New Type of Matter

1. Charles Seife*
1. A special Web feature on this year's physics laureates, including research, news, and commentary from the pages of Science, can be found at http://www.sciencemag.org/feature/data/nobelprize/2001/physics.shtml

Three new laureates per prize—the maximum number Nobel rules allow—gain recognition for fundamental advances in their fields

Wolfgang Ketterle, Eric Cornell, and Carl Wieman have gotten a warm reception for their chilly work: They have won the 2001 Nobel Prize in physics for creating the first Bose-Einstein condensates (BECs) in gases of rubidium, sodium, and other alkali metals.

“It's very well deserved,” says Claude Cohen-Tannoudji, a physicist at the école Normale Supérieure in Paris. “There [are] a lot of new directions being explored” because of BECs, he adds.

By cooling gases to a few billionths of a degree above absolute zero and coaxing them into forming a new state of matter, the three laureates verified a prediction made by Albert Einstein 70 years earlier. Einstein, in turn, took his cue from physicist S. N. Bose, who, in the mid-1920s, investigated the properties of particles that have integer spin—now termed “bosons.” Bosons, which include certain atoms, behave differently from their opposite numbers, fermions, which have half-integer spins. Fermions tend to avoid one another; that is why you can fit only a certain number of electrons, which are fermions, into each atomic shell. Bosons, on the other hand, have no such restrictions, so many of them can occupy the same atomic state at the same time.

Einstein claimed that when cooled enough, bosons in a gas would stop jittering about and settle down into the lowest energy state, or ground state. Thanks to their sociable nature, thousands of bosons could all be in the ground state, forming, in a sense, one large “superboson”: a BEC. BECs are playgrounds for bizarre physics. You can manipulate a BEC to create a very fine interference pattern, slow light down to a crawl within it (Science, 27 July, p. 663), or use it as an almost macroscopic testing ground for quantum mechanics. “We brought it to an almost human scale,” says Wieman. “We can poke it and prod it and look at this stuff in a way no one has been able to before.”

For decades, researchers tried to inveigle matter into becoming a BEC, without success. Then, in 1995, Cornell and Wieman, physicists at the University of Colorado, Boulder, used a combination of optical and magnetic trapping techniques to bully about 2000 cooled rubidium atoms into forming a BEC. Shortly thereafter, Wolfgang Ketterle of the Massachusetts Institute of Technology created a considerably bigger BEC cloud out of sodium atoms. Those achievements set off a flurry of experiments in which teams watched BECs interfere with themselves, used them to create “atomic lasers,” and watched as vortices formed and dissipated within the BECs. Researchers have also added new atoms to the roster of BEC-producing gases, including isotopes of hydrogen, lithium, and most recently potassium (http://www.sciencexpress.org/). “We've been surprised to see the explosive growth of the field,” Ketterle says. “We thought it would be neat, but it has had an enormous impact on atomic physics.”

The prize, split evenly among the three winners, comes as no surprise to the physics community. In 1997, Cohen-Tannoudji, along with physicists Steven Chu of Stanford University and William Phillips of the National Institute of Standards and Technology in Gaithersburg, Maryland, won the Nobel Prize in physics for developing the cooling techniques that enabled physicists to make BECs. That prize was widely seen as an early acknowledgment of the importance of BEC research. Now the other shoe has dropped, and the physicists who created the first BECs can bask in glory that is far more than cold comfort.

12. NOBEL PRIZES

# Science Awards Pack a Full House of Winners

1. Robert F. Service

Three new laureates per prize—the maximum number Nobel rules allow—gain recognition for fundamental advances in their fields

Pioneers of chemical handedness received their own round of applause from Stockholm last week. A pair of U.S. researchers—William Knowles, now retired from Monsanto in St. Louis, Missouri, and K. Barry Sharpless of the Scripps Research Institute in La Jolla, California—along with Japanese chemist Ryoji Noyori of Nagoya University won this year's Nobel Prize in chemistry for creating catalysts that can produce a particular geometric form of a compound without creating its mirror-image partner. The work has proved vital for the production of everything from pharmaceuticals and flavorings to advanced materials and insecticides.

George Whitesides, an organic chemist at Harvard University, says the Nobel committee likely had a difficult choice because other researchers, including French chemist Henri Kagan, also played important roles in developing such catalysts. Kagan, Noyori, and Sharpless received this year's Wolf Prize in chemistry, often seen as a predecessor to the Nobel. Nevertheless, Whitesides says he is “delighted” that Sharpless, Noyori, and Knowles were honored. “This combination of guys put together what has become a dominant theme in organic synthesis. It has motivated the chemical community for a number of years.”

Chemists have known since the 1870s that molecules can come in more than one mirror-image form. This property, known as chirality, is widespread in biology: DNA, proteins, and sugars all boast mirror twins. In some cases this slight structural difference can lead to dramatic consequences. When the drug thalidomide was given to pregnant women in the 1960s to prevent nausea, one of its mirror-image forms caused birth defects in thousands of children.

At the time, most techniques for making handed molecules produced only mixtures of chiral forms. Attempts to purify out the desired compound were inefficient, costly, and wasteful, says Harvard organic chemist Stuart Schreiber. Knowles set out to make a catalyst specific to one side of the mirror. He worked on one that added hydrogen atoms to molecules harboring pairs of carbon atoms. The carbon pairs sit in a flat plane, and the hydrogens poke out either above or below—which way determines the handedness of the final compound. Other early catalysts attached hydrogens to either side indiscriminately. But in 1968 Knowles came up with a novel version which, he says, “was shaped so that it could only come in on one side.” He soon used this strategy to devise an industrial process to make L-dopa, an amino acid useful in treating Parkinson's disease.

Noyori later expanded on Knowles's early work to create more broadly useful hydrogen-adding chiral catalysts that are still widely used in industry today. For their work, the Royal Swedish Academy of Sciences awarded them half of this year's chemistry prize. Sharpless earned the other half for creating chiral catalysts that add oxygen to precursor molecules. That has proven to be an even more versatile tool, says Whitesides, because it creates chiral building blocks that can be easily modified further to make a wide range of materials and drugs. “Chemists used those to solve a bunch of synthetic problems that had previously been insoluble,” he says.

13. NOBEL PRIZE IN ECONOMICS

# Knowledge Is Power

1. Barry Cipra

Three new laureates per prize—the maximum number Nobel rules allow—gain recognition for fundamental advances in their fields

In classic economic theory, there are no secrets; everybody knows what everybody else knows. The real world, of course, doesn't work that way. In most market settings, one side knows more than the other: a car dealer, say, who knows he's unloading a lemon, or a job applicant who knows how hard a worker she really is. Economists began exploring the implications of “asymmetric information” in the 1970s. The fruits of their labors have now netted three of them a Nobel Prize.

The latest laureates are George Akerlof of the University of California, Berkeley, Michael Spence of Stanford University, and Joseph Stiglitz of Columbia University. Their findings undercut many once-cherished assumptions of economics, explains Andrew Weiss of Boston University, including the mantric “law” that supply equals demand, which may not hold when information is unequally held. “This is absolutely revolutionary work,” Weiss says. “It changes the way we teach economics.”

Akerlof's groundbreaking work, a 1970 essay titled “The Market for Lemons,” which the Nobel Prize committee calls “the single most important study in the literature on economics of information,” showed how asymmetric information can create a market in which supply and demand are out of whack. If buyers can't tell the difference between a peach and a lemon (until it's too late), they won't be willing to pay what a peach is actually worth. As a result, the sellers of peaches retreat from the market, leaving only lemons—even though demand for peaches is still strong. A recent, costly example involves the information technology (IT) sector of the economy itself: When prominent but unprofitable IT “lemons” began going belly up, investors realized they had no way of distinguishing them (in advance) from profitable ventures and pulled out altogether.

One way to cope with asymmetric information is for informed parties to “signal” their information. In his 1972 dissertation, Spence analyzed how job applicants use educational attainment to signal their potential productivity. Racking up an advanced degree or a grade-inflated GPA may not mean you know anything useful, but it presumably says something about your work ethic. Spence found that informed parties can be driven to overinvest in such signals—be it a person who gets more education than he or she needs or a company that posts a big dividend instead of expanding, just to show investors it's making a profit. “Mike's fundamental insight lets us understand a huge range of real world phenomena,” says Stanford colleague John Roberts. “Such insights come very, very rarely.”

Uninformed parties can also elicit information from the other side. In a 1976 paper co-authored with Michael Rothschild of Princeton University, Stiglitz showed how insurance companies do this. By offering, for example, high-premium/low-deductible vs. low-premium/ high-deductible car insurance, companies in effect get customers to declare how risky or safe a driver they are.

Revolutionary as it was, asymmetric information is now widely accepted. “It's part of the canon of economic theory,” Rothschild says. What remains to be seen is whether the supply of new ideas can meet the demand for solutions to economics problems.

14. INFECTIOUS DISEASES

# Hard-Won Advances Spark Excitement About Hepatitis C

1. Charlene Crabb*
1. Charlene Crabb writes about science from Paris.

Years of research are paying off with a new mouse model and insights into natural immunity, raising hopes about drugs or vaccines to fight the epidemic

PARIS—In the mid-1970s, Harvey Alter of the U.S. National Institutes of Health sent out an SOS. A new form of hepatitis was attacking the livers of some blood-transfusion recipients, and Alter wanted help tracking down the culprit. Not until 1989 did Michael Houghton and his team at Chiron Corp. in Emeryville, California, identify the elusive pathogen—hepatitis C virus (HCV). Blood centers began screening for the virus the following year, stanching the flow of HCV-contaminated blood into the blood supply.

But scientists soon realized they were dealing with a viral time bomb that they had few tools with which to disarm. Spread primarily by direct contact with human blood, HCV had already infected millions of people through transfusions or unsafe medical practices. HCV continues to infiltrate developed countries, mainly by needle-sharing drug users. And in developing countries, medical practitioners reusing inadequately sterilized needles inadvertently inoculate millions of patients with the pathogen.

At least 3 million to 4 million people worldwide are infected each year, according to the World Health Organization, compared to an estimated 5.6 million new infections of the virus that causes AIDS. Once infected, about 20% of people clear the virus from their bloodstream, but the rest will harbor HCV the rest of their lives. Of those chronic cases, many may never have symptoms, but 10% to 20% eventually develop liver-destroying cirrhosis or cancer.

As the extent of the HCV epidemic hit home, dozens of research teams began trying to find ways to prevent infection or more effectively treat it. But the scientists immediately ran into a brick wall: For reasons still baffling, the virus stubbornly refuses to grow in the lab, greatly handicapping research efforts. Now, a dozen years after unmasking the pathogen, there is still no vaccine or specific antiviral drug. The best available treatment combines interferon, which modulates the immune system, with ribavirin, a nonspecific antiviral drug. But the regimen succeeds in only 30% to 50% of cases, and side effects are significant.

Last month, some 1000 hepatitis C researchers met here at an annual conference to discuss the headway they have made recently in understanding and taming the virus—including a new animal model—as well as some letdowns.

At a meeting 2 years ago, hopes were raised that the brick wall was beginning to crumble when scientists announced that they had devised a potential way to grow HCV in tissue cultures (Science, 2 July 1999, p. 110). Ralf Bartenschlager and his team at the Institute for Virology at Johannes Gutenberg University in Mainz, Germany, built truncated versions of the virus's genetic material, called replicons. Inserted into human cells grown in tissue culture, the replicons began to churn out proteins the real HCV uses to make copies of itself. About a year later, the Mainz team as well as researchers led by Charles Rice, then at Washington University School of Medicine in St. Louis, Missouri, announced separately that they had each improved on the replicon by identifying mutations in HCVgenes that increased replication (Science, 8 December 2000, p. 1870). Since then, several scientists have been adding the complete complement of HCV genes to the truncated replicons, hoping to produce intact virus.

But at last month's meeting, Bartenschlager's team announced that although they successfully made an entire HCV genome that replicates efficiently in tissue culture, the proteins it produces fail to knit themselves together into virus particles. “It is disappointing,” says Bartenschlager. “What's more disappointing is we really don't know why. Maybe there is a host cell factor required for proper interaction of the proteins, and this factor is missing in the cell culture. But that's just one possibility.”

## Supermice

While researchers continue trying to grow an intact virus in cells, Norman Kneteman and a team of surgeons and scientists at the University of Alberta in Edmonton, Canada, have devised a way to grow HCV in mice: by giving them human livers. The chimeric mice are a handier and less expensive in vivo model than chimpanzees, the only animal besides humans that HCV infects. “They will revolutionize the testing of antivirals,” predicts Christopher Richardson of the Ontario Cancer Institute in Toronto, who is already using the mice to test a potential gene therapy for HCV infection.

These are no ordinary mice, even before they receive grafts of human liver. Kneteman's group spent 3 years selectively breeding them from two types of genetically engineered mice: mice with severe combined immunodeficiency (SCID) and mice carrying a sequence of genes, called Alb-uPA, that disrupts normal liver cell function. The Alb-uPA mice have sickly, white livers. But, thanks to the unique ability of mammalian livers to regenerate, faulty livers are constantly being stimulated to grow new cells. So, the offspring of the Alb-uPA and SCID mice have livers that work in overdrive to grow new cells and immune systems that are unable to recognize foreign tissue.

And that is just what the researchers needed for successful transplants of human tissue, as they reported in the 1 August issue of Nature Medicine. With a technique honed to a 5-minute procedure, the surgeons transfer about a million human hepatocytes, taken from either fresh or frozen liver tissue, into the spleen of a 10- to 14-day-old SCID/Alb-uPA hybrid. After 6 weeks, the rodent liver boasts red nodules of functioning human cells. Mice that receive a copy of the Alb-uPA gene sequence from each parent grow enough human liver to develop a chronic hepatitis C infection when inoculated with the virus.

Kneteman's group is already using the homozygous SCID/Alb-uPA hybrids in a number of experiments. At the meeting, the team reported that HCV-infected mice treated with interferon for 2 weeks had lower amounts of HCV in their blood than untreated mice. Scientists think interferon works by both revving up the immune system and directly disarming HCV. Because the mice lack key immune system defenses, having fewer viral particles indicates that interferon does indeed exert an antiviral effect. Next, the team plans to expand the study to the widely used combination of interferon and ribavirin, the newer pegylated interferon, and small antiviral molecules.

But the new SCID/Alb-uPA mice have drawbacks. “They are tricky to propagate,” explains Rice, now at Rockefeller University in New York City. “And anything that involves getting very fresh and, hopefully, normal liver tissue is challenging.”

Scientists won't have to build their own mice, however. Kneteman and colleagues have applied for a patent and formed a company, KMT Hepatech, to expand their breeding colony of homozygous mice and make them available to researchers. For now, to ensure quality control, all studies with the mice will be done in the Edmonton lab, Kneteman says. Outside scientists doing basic research on HCV can use the mice in exchange for co-authorship of published papers resulting from the work; studies of potentially commercial ideas or products would be covered by an upfront agreement to share research costs and resulting revenue. “Our goal is to make the biggest impact on the disease,” says Kneteman. “We don't want to limit the availability of these animals.”

## Vaccine hopes

Although the lack of a small-animal model has been a big handicap for efforts to develop a vaccine against HCV, researchers have also been discouraged by hints that the immune system doesn't seem to remember earlier encounters with the virus. This could be a huge drawback, because vaccines rely on immunologic memory: the ability to recognize an earlier invader. At the meeting, however, several groups reported that human and chimpanzee immune systems do indeed remember HCV and can thus mount quick and robust counterattacks.

In the early 1990s, scientists noted that chimps that had overcome an initial acute HCV infection would develop acute infections again whenever they were reinoculated. The reinfections suggested that T cells and other components of the immune system failed to recognize HCV even though they had battled it before. Those studies “really boded poorly for HCV vaccine development,” says Christopher Walker of the Children's Research Institute at Ohio State University, Columbus.

Walker is one of several researchers who recently repeated the “rechallenge” experiments on chimps, using more sensitive techniques for measuring the level of cellular immune response. Examining the liver and blood of chimps that had cleared an HCV infection 4 to 5 years earlier, Walker's group found that large numbers of CD4 and CD8 T cells were still primed to attack the virus. When reinoculated with HCV, the chimps developed less severe infections compared with their first bout with the virus. The number of viral particles in the blood at the peak of infection was 100 times lower, and the virus disappeared from the blood in an average of 14 days rather than 4 months.

Human immune systems are similarly primed, other scientists reported. Georg Lauer of Massachusetts General Hospital in Boston described two patients who had shrugged off HCV infections at least 2 years earlier but still had a “huge” number of CD8 T cells—at least 5%—preprogrammed to recognize HCV. The cases are “rare extremes,” says Lauer, but they could provide insights into mechanisms for controlling HCV or developing long-term immune responses to it.

Another team, led by David Thomas of Johns Hopkins University School of Medicine in Baltimore, reported that humans have shorter and milder acute infections on subsequent exposure to the virus, much as chimpanzees do. In a 2-year study, Thomas and colleagues tracked HCV infections in more than 250 injection drug users who came in for checkups every 6 months. During the study, some individuals became infected with HCV for the first time, and others became reinfected. When the researchers compared the two groups, they found that those with first-time infections, as opposed to reinfections, were twice as likely to have viral particles in their blood for at least two consecutive visits and the number of particles was 100 times more at peak infection.

Finding natural immunity, say Thomas and others, suggests that a therapeutic vaccine may be possible. Such a vaccine may not prevent an acute infection, but it might prime the immune system to clear out the virus quickly and prevent a life-threatening chronic infection. A vaccine might also benefit people with chronic infections by eliminating the virus or keeping the viral load in check so that cirrhosis or cancer would be less likely to develop.

Pharmaceutical companies are already working on such vaccines. Innogenetics NV in Ghent, Belgium, for instance, is now conducting a phase II clinical trial for a therapeutic vaccine based on recombinant E1, a viral glycoprotein found on the surface of HCV. Chronically infected patients produce few or no E1 antibodies, whereas patients who shrug off the virus make large quantities of E1 antibodies, often in combination with a strong T cell response. Similarly, Chiron is gearing up for phase I trials of two potential vaccines. One is based on E1 and E2, another viral surface glycoprotein. Chiron's second candidate is based on the HCV core protein, which surrounds the genetic material of the virus. Chiron's Houghton envisions using the vaccines “alone or in combination.”

“The mood among researchers has really changed,” says Houghton. “There's now more optimism for vaccine development. More and more of us think we may be able to protect the majority of individuals from chronic HCV infection.”

15. ALZHEIMER'S DISEASE

# Bad for the Heart, Bad For the Mind?

1. Jean Marx

High cholesterol levels may foster the brain degeneration of Alzheimer's, raising the possibility that cholesterol-lowering drugs will protect against the disease

Cholesterol has a bad reputation, and justifiably so, because elevated levels increase the risk of cardiovascular diseases such as heart attack and stroke. But it may deserve even more condemnation. Accumulating evidence suggests that high cholesterol levels contribute to Alzheimer's disease as well.

Researchers have built a case over the past few years that a small protein called β amyloid (Aβ) causes the brain degeneration of patients with Alzheimer's. But they have also suspected that Aβ has accomplices. High blood cholesterol levels may be one of these malefactors, possibly because they aid and abet Aβ production, several teams now suggest. “It looks like cholesterol is involved [in Alzheimer's] in some fashion,” says Neil Buckholtz, who oversees Alzheimer's research at the National Institute on Aging (NIA) in Bethesda, Maryland.

If so, cholesterol may provide “an accessible and straightforward target for decreasing Aβ and reducing Alzheimer's risk,” says Joseph Buxbaum of Mount Sinai School of Medicine in New York City. Indeed, NIA has just announced that it will expand its ongoing Alzheimer's Disease Cooperative Study (ADCS) to include a trial aimed at determining whether the cholesterol-lowering drugs called statins slow the progression of mild to moderate Alzheimer's. A similar trial is already under way at the Sun Health Research Institute in Sun City, Arizona, and researchers in Europe are also looking at statins' effects on Alzheimer's.

One of the first clues that cholesterol might be involved in the disease came about 10 years ago from Larry Sparks and his colleagues at the University of Kentucky Chandler Medical Center in Lexington. Sparks, then a forensic pathologist, was often called upon to perform autopsies on people who had died unexpectedly. He noticed that about 70% of individuals who had succumbed to heart disease also had amyloid-containing plaques—one of the defining features of Alzheimer's pathology—in their brains. People of about the same age who had died of other causes were much less likely to have plaque-ridden brains, leading Sparks to suspect a link between high cholesterol levels and Alzheimer's.

That was far from proof, however; impaired blood flow to the brain or other factors might explain why heart disease victims developed so many plaques. But 2 years later, the cholesterol connection got a boost when a genetic linkage study by Allen Roses, Warren Strittmatter, and their colleagues at Duke University School of Medicine in Durham, North Carolina, identified a particular variant of the gene that encodes the cholesterol-carrying apolipoprotein E as an Alzheimer's risk factor. Exactly how that variant, known as ApoE4, predisposes someone to the disease is not well understood, but since then several studies have also pointed to a link between cholesterol and Alzheimer's.

Two of the most intriguing are epidemiological studies that appeared last year. In these studies, researchers examined whether statins, which are taken by millions of people, influence the risk of getting Alzheimer's. “Since patients have been taking statins for years now, we thought we might be able to see an effect if it were present,” says Benjamin Wolozin of Loyola University Medical Center in Maywood, Illinois, who led one of the studies.

The studies used different statistical designs and focused on different populations. Wolozin's team gathered data from patient records in two hospitals in Illinois and one in Arizona. Meanwhile, Herschel Jick of Boston University School of Medicine, David Drachman of the University of Massachusetts Medical School in Worcester, and their colleagues obtained data from the U.K.'s General Practice Research Database, a compendium of demographic and medical information from 3 million U.K. residents. Even so, the teams' conclusions were strikingly similar, Wolozin says: “The prevalence of Alzheimer's disease in people taking statins was about 70% lower” than in controls.

But Drachman is not sure that the effect was due primarily to the statins' cholesterol-lowering action. His team looked at the effects of other cholesterol-lowering drugs—and they did not reduce the risk of dementia. “The issue,” Drachman says, “is what do the statins do that the other drugs don't do?” As one possibility, he points out that statins increase the activity of an enzyme called nitric oxide synthase in the blood vessel linings and decrease the activity of another protein called endothelin-1. As a result, they improve the function of small blood vessels and increase blood flow to the brain.

Other research with both cultured cells and animals, however, supports the premise that statins protect against Alzheimer's because they block cholesterol synthesis. In the mid-1990s, Sparks and his colleagues found that rabbits fed a high-cholesterol diet developed plaques and other signs of Alzheimer's pathology in their brains. The next question, says Sparks, now at the Sun Health Research Institute, was “Can we make this go away?” They could. Switching the rabbits to their normal low-cholesterol diet reduced the number of plaques. And even rabbits that gorged on a high-cholesterol diet developed far fewer plaques if they were also given probucol, a nonstatin cholesterol-lowering drug.

Clues to just how excess cholesterol might spur amyloid plaque formation came a few years later in studies of brain neurons in culture. In the past few years, several teams have shown that adding cholesterol to neurons makes them churn out more of the plaques' active ingredient: Aβ. Conversely, statins, as well as chemicals that extract cholesterol from the cell membrane, decrease formation of the peptide.

Neurons in culture are a far cry from those in the human brain, but other work shows that cholesterol-lowering drugs also minimize Aβ production in the brains of living animals. Sparks previously demonstrated this in rabbits, and earlier this year, a team led by Konrad Beyreuther and Tobias Hartmann of the University of Heidelberg, Germany, found that simvastatin lowers Aβ levels in the brains and cerebral spinal cords of guinea pigs. And in as-yet-unpublished work, Larry Refolo, Karen Duff, and their colleagues at the Nathan S. Kline Institute for Psychiatric Research in Orangeburg, New Jersey, showed that an experimental cholesterol-lowering drug reduces production of the peptide in the brains of genetically engineered mice that develop plaques much like those in human Alzheimer's.

An indication that something similar might happen in humans comes from Mount Sinai's Buxbaum, Lawrence Friedhoff of Andrx Corp. in Hackensack, New Jersey, and their colleagues, who found that clinical doses of lovastatin reduce Aβ levels in the blood of human patients in a dose-dependent manner. The researchers were unable to see how the drug affects Aβ production in the patients' brains, but Buxbaum says that because lovastatin penetrates the brain, it's possible that it lowers Aβ production there, too.

Work with cultured cells is providing clues to how cholesterol-lowering drugs reduce production of Aβ. Cells produce the peptide by clipping it out of a larger protein called APP (for β-amyloid precursor protein), with the aid of two enzymes, known as the β- and γ-secretases. But APP is also cut by a third enzyme, the so-called α-secretase. Because this enzyme breaks APP within the Aβ segment, it prevents production of the neurotoxic peptide. Results reported in two papers in the 8 May issue of the Proceedings of the National Academy of Sciences (PNAS)—one from Beyreuther, Hartmann, and their colleagues and the other from a team led by Falk Fahrenholz of Johannes Gutenberg University in Mainz, Germany—suggest that cholesterol-lowering treatments inhibit Aβ formation by shifting the balance of activities of these enzymes to favor the α-secretase.

Fahrenholz and his colleagues found that the statin they used increased production of an α-secretase called ADAM10, but he notes that other effects on membrane biochemistry may have also contributed to the shift. And the Heidelberg team's PNAS paper suggests that cholesterol reduction decreases γ-secretase activity, although Hartmann says that more recent work suggests that it decreases β-secretase activity as well. The exact cause of the altered enzyme function is unclear, however. As Duff points out, “we know cholesterol affects amyloid load, but we don't know how.”

One possibility is that changes in cholesterol content affect the activity of the APP-cleaving enzymes by altering their physiological milieu. Lipids and proteins aren't evenly distributed in cell membranes. For example, there's some evidence that APP and the β- and γ-secretases are located together in cholesterol-rich areas of the membrane known as “rafts.” So removing cholesterol may disrupt that association, while at the same time making it easier for α-secretase, which prefers an environment rich in a different lipid called sphingomyelin, to access APP.

Yet another clue about cholesterol's relationship to Aβ comes from Dora Kovacs, Luigi Puglielli, and their colleagues at Massachusetts General Hospital in Boston, who made an intriguing observation about a possible role for an enzyme called acyl-coenzyme A: cholesterol acetyltransferase (ACAT). When cholesterol levels in the membrane get too high, ACAT removes the excess and adds on an acyl group, thus forming a cholesterol ester, which is stored in granules inside the cell.

Working with cells with mutations that either increase or decrease the formation of cholesterol esters, the Kovacs team showed that Aβ production correlates not with total cellular cholesterol but with cholesterol ester levels. Consistent with that, the Mass General workers also found that compounds that inhibit ACAT, and thus cholesterol esterification, also inhibit Aβ release.

The researchers, who describe their results in the October issue of Nature Cell Biology, do not yet know how high cholesterol ester levels boost Aβ production. In fact, Kovacs says, her team was “definitely surprised” by the result. It doesn't easily fit in with the fact that APP and the three secretases are found in the cell membrane, where it would be difficult for them to interact with granules. More work will clearly be needed to sort this out. But meanwhile, Kovacs says, “the important thing is that this means that inhibitors of ACAT could theoretically be used as inhibitors of Aβ production” and thus as Alzheimer's drugs.

Wolozin, for one, welcomes that idea, calling the Kovacs team's findings “really important.” He and others point out that statins, which block a very early step in cholesterol synthesis, may not be ideal for long-term use for Alzheimer's prevention. The drugs have been associated with sometimes dangerous side effects. A few months ago, the Bayer Group of Leverkusen, Germany, recalled its statin, known as cerivastatin or Baycol, because several people died from a condition called rhabdomyolysis, in which the muscles break down. That problem may have been linked to the simultaneous use of another, nonstatin cholesterol-lowering drug called gemfibrozil, but statins alone can also cause liver damage in some people.

Still, researchers are eager to see how the drugs fare in the trials now under way or planned. It will be at least a year before results of the Arizona trial are in, and the ADCS study hasn't begun yet. Both trials will be conducted on patients who have early symptoms of Alzheimer's and therefore have already suffered considerable brain degeneration. This may make it more difficult to see treatment benefits.

That should not be a problem, however, with the Prospective Study of Pravastatin in Elderly at Risk (PROSPER) now being conducted on more than 5000 people in Scotland, Ireland, and the Netherlands. In addition to seeing how the statin affects the cardiovascular health of these people, researchers will also follow its effects on their cognitive function. Results are expected next year. If the statins or other cholesterol-lowering drugs do benefit patients, we can conclude that lowering cholesterol helps the mind as well as the heart.