# News this Week

Science  15 Aug 2008:
Vol. 321, Issue 5891, pp. 898
1. ANTHRAX INVESTIGATION

# Full-Genome Sequencing Paved the Way From Spores to a Suspect

1. Martin Enserink*
1. With reporting by Yudhijit Bhattacharjee.

The scientific evidence against Bruce Ivins, the 62-year-old Army scientist who killed himself while about to be indicted for the anthrax murders, is finally emerging. Last week, the Federal Bureau of Investigation (FBI) laid some of its cards on the table. One key document, scientists say, now enables a reconstruction of the trail that led the FBI from the deadly letters back to Ivins's lab at the U.S. Army Medical Research Institute of Infectious Diseases (USAMRIID) in Fort Detrick, Maryland.

The investigation relied heavily on outside labs such as The Institute for Genomic Research (TIGR) in Rockville, Maryland, which sequenced a large number of anthrax samples; it also required the development of new genetic tests. Although none of the steps was revolutionary or particularly inventive, researchers say, combining them to solve a criminal case was. Surprisingly, many past speculations on the forensic science were wrong on one point: Sophisticated fingerprinting techniques for Bacillus anthracis developed at Northern Arizona University (NAU) in Flagstaff, widely rumored to be crucial, didn't play a significant role.

Scientists say they need many more details to decide the merits of the case against Ivins. But despite the bureau's widely ridiculed mistakes—including an early focus on Ivins's former colleague Steven Hatfill—“the scientific evidence is probably really strong,” says Steven Salzberg, a former TIGR researcher now at the University of Maryland (UMD), College Park. “They've got some very good people,” Salzberg says. “The impression that they're not good may just come from their style. They never tell you anything.”

The main document unsealed last week is an October 2007 affidavit by Thomas Dellafera, a postal inspector. Filed in support of a warrant to search Ivins's home, cars, and a safety box, the 25 pages of text didn't spell out the details of the evidence. But a close reading of the four paragraphs about the FBI's genetic analysis helps clarify how the bureau approached the problem, says microbiologist Jeffrey Miller of the University of California, Los Angeles.

The key to understanding the investigation is that the anthrax used in the attacks didn't have a single, uniform genetic makeup, a source close to the investigation says. Each of the envelopes likely contained many billions of spores; within such a population, there are always subpopulations of cells bearing mutations that set them apart from the majority. The same minorities would presumably have been present in the “mother stock” of anthrax from which the spores were prepared.

However, standard sequencing—which would require the DNA from thousands of spores—would have resulted in a “consensus sequence” for the spores, in which such rare mutations were simply drowned out. To find them, researchers used a different technique: They grew spores from the envelopes on petri dishes, generating hundreds or even thousands of colonies per dish, each the progeny of a single spore. They then searched for colonies that looked different from the majority; the affidavit mentions variations in “shape, color, texture.” (Those colonies might have been rough instead of smooth, or much smaller than most, Miller says.) Next, they set out to find the mutations that made those colonies different.

To do that, the FBI used a brute-force approach: It had the entire genomes of the bacteria in the minority sequenced. TIGR—which merged into the J. Craig Venter Institute in 2006—sequenced “probably somewhere between 10 and 20” such genomes in the years after the attacks, Salzberg says. TIGR could not handle live anthrax cells itself; the FBI gave the lab purified DNA produced by Paul Keim's lab at NAU, Salzberg says. Claire Fraser-Liggett, who led TIGR at the time and is now also at UMD, declines to discuss details of the investigation. But two other sources confirm TIGR's role.

Comparing the sequence of the variant colonies to an original B. anthracis strain called Ames, widely used in research, identified a number of mutations, says Salzberg; they included single-nucleotide polymorphisms, a change of a single base pair, and tandem repeats, in which a short piece of DNA is repeated a variable number of times.

The FBI then had scientists at other labs develop tests that allowed them to screen any anthrax sample for four of these mutations. Such assays “are very easy to design,” for instance, using a polymerase chain reaction-based strategy, says evolutionary biologist Richard Lenski of Michigan State University in East Lansing; molecular biology labs do it all the time.

Armed with the four tests, the FBI examined more than 1000 anthrax isolates, collected from 16 labs that had the Ames strain in the United States and several more in Canada, Sweden, and the United Kingdom. In only eight of those samples, they found all four mutations seen in the envelope samples; and each of these eight, the affidavit says, was “directly related” to a “large flask” of spores, identified as RMR-1029, which Ivins had created in 1997 and of which he was the “sole custodian.”

That still leaves many questions open, researchers say. One thing that needs to be explained, says Miller, is whether the eight isolates that were “directly related” to RMR-1029 were all found at USAMRIID, or whether some came from other laboratories. In the latter case, it's unclear why the FBI ruled out those labs as the potential origin. (One clue that the affidavit offers is that USAMRIID is the only lab in Maryland or Virginia, the states where the particular envelopes used in the attacks were sold.)

It's also unclear how many of the 1000 samples had fewer than four, but more than zero, of the mutations. “If a whole bunch of them had two or three,” that would increase the odds that the perfect match at USAMRIID was just a false positive, Lenski says. Another key question, he adds: Where in the anthrax genome did the four mutations occur? If they were in hypervariable regions, that would also probably make the case against Ivins weaker.

Whether the analysis would hold up in court seemed to be front and center in the FBI's thinking, says Salzberg. For instance, when researchers from TIGR and NAU published a comparison of two anthrax strains in Science in 2002 (14 June 2002, p. 2028), a top FBI researcher named Bruce Budowle encouraged them to include a statistical analysis to estimate the data's accuracy, Salzberg says. “Budowle felt it would be useful to have it all go through peer review, in case it went to court,” he says.

The FBI has invested heavily in microbial forensic expertise since 2001, and Budowle has co-authored many papers on the topic. But the bureau farmed out much of the scientific bench work, in part because the Marine Corps doesn't allow bioweapons agents at its base in Quantico, Virginia, where the FBI Laboratory is located. The work was “highly compartmentalized,” says a source close to the investigation: Most labs didn't know exactly what the others were doing.

The affidavit is very unclear about whether the spore preparations might have undergone physical or chemical treatments to make them disperse more easily—still a point of major confusion, says Barbara Hatch Rosenberg, a bioweapons specialist at Purchase College in New York. Scientists at the Armed Forces Institute of Pathology reported in October 2001 that the spores sent to U.S. Senator Tom Daschle's office had been mixed with silica to make them more easily dispersible. However, in congressional briefings and in a paper published in the August 2006 issue of Applied and Environmental Microbiology, FBI officials described the powder as a simple spore preparation without additives.

The affidavit reports that there was “an elemental signature of Silicon within the spores” in all four letters that were recovered. This silicon signature is later cited as part of the evidence linking the mailed anthrax to the flask of spores that Ivins had access to. But what the silicon was for, or whether other samples were tested for the signature, remains unclear.

Science aside, the affidavit relies heavily on circumstantial evidence. For instance, it notes unexplained spikes in Ivins's nighttime lab activity right before the two waves of letters were sent. It also claims that he tried to mislead investigators to hide his involvement. In April 2002, he submitted samples from his lab that tested negative for the four mutations, according to the affidavit; but on 7 April 2004, an FBI agent seized the RMR-1029 flask, which tested positive for all four. Ivins insisted he had given agents RMR-1029 the first time around, however.

One of the weak points in the affidavit is Ivins's motive, says Gregory Koblentz, a biodefense specialist at George Mason University in Fairfax, Virginia. The FBI suggests that Ivins was afraid of losing his job if the government ended a project he was working on that was trying to solve regulatory issues around the so-called AVA anthrax vaccine. It “seems a bit of a stretch” that Ivins would have thought his job hinged on that project, says Koblentz. His group “would have had plenty of other anthrax vaccine-related work to keep them busy.” A glaring omission, meanwhile, is any evidence placing Ivins in Princeton, New Jersey, on any of the days the envelopes could have been mailed from there.

A spokesperson for the FBI's laboratory declined a request to interview Budowle and referred scientific questions to the FBI's Washington, D.C., field office. “In the near future the FBI will determine the best way to address the science involved in the anthrax case,” the spokesperson e-mailed Science. Many suspect that with so many burning questions, a full account of the evidence—including the scientific details—is now just a matter of time.

2. CLIMATE PREDICTION

# Seasonal-Climate Forecasts Improving Ever So Slowly

1. Richard A. Kerr

Farmers, ski-resort operators, and heating-oil suppliers would very much like to know what the coming winter will be like. If a strong El Niño were brewing in the tropical Pacific, at least some of them would be in luck. The official United States winter forecast could warn them, with considerable reliability, that the Southeast and the Gulf Coast will be cooler and wetter than normal. But without an El Niño or its counterpart, La Niña, next winter's weather is pretty much anybody's guess.

Of the dozens of forecasting techniques proffered by government, academic, and private-sector climatologists, all but two are virtually worthless, according to a new study. “There are seasons, places, and situations in which skill is very, very good,” says climatologist and study co-author Robert Livezey, recently retired from the National Weather Service (NWS). But even many people in the field “don't appreciate how little there is to work with. There is really no evidence here that there are any other silver bullets” waiting to be found.

Since 1946, NWS forecasters have been trying to forecast the average temperature and precipitation across the lower 48 states a month ahead, and more recently season by season up to a year ahead. At NWS's Climate Prediction Center (CPC) in Camp Springs, Maryland, where Livezey oversaw seasonal forecasting in the late 1990s, the trick has generally been to identify some element of recent or current climate—say, the presence of El Niño—that can influence future climate. If they couldn't find one, researchers could fashion a forecast “tool”—such as a collection of past time periods when the climate system resembled the current situation—that when tested on past seasons gave some inkling of future seasons. They would then subjectively choose which techniques to combine and how to combine them in order to predict whether temperature and precipitation would be above, near, or below normal in some 3-month period in a particular region.

The CPC approach has shown very modest though increasing skill at CPC, Livezey and climatologist Marina Timofeyeva of NWS in Silver Spring, Maryland, report in the June issue of the Bulletin of the American Meteorological Society. They worked up a scorecard for CPC forecasts made from 1994 to 2004, comparing the success rates for different seasons, regions, and periods when a strong El Niño or La Niña was present or absent.

About the only time forecasters had any success predicting precipitation was for winters with an El Niño or a La Niña, Livezey and Timofeyeva found. Using a scale in which mere chance is 0% and perfection is 100%, in those winters they estimate “unprecedented” skill—50% to more than 85%—along the southern tier states and up the West Coast about half a year into the future. Even so, the overall skill score for precipitation was just 3%.

Temperature forecasts fared better, with an overall skill score of 13%, up from a score of 8% for the previous decade. El Niño and La Niña helped out again during winter, raising skill to more than 85% across much of the eastern United States out to more than 8 months. But CPC also had substantial success predicting temperature out to a year in the American West outside of El Niño-La Niña years, thanks to the long-term greenhouse warming trend picked up by one of the forecast tools.

Because a strong El Niño or La Niña shows up only every few years, his results paint “a kind of discouraging picture” of seasonal forecasting, Livezey says: “You can probably find dozens of forecast [techniques] people use to give themselves an edge. Almost all of that is mumbo jumbo.” CPC forecasters have done well to make their forecasts more objective in recent years, Livezey and Timofeyeva write; CPC should weed out remaining weak forecast tools and focus future research on computer model forecasting of climate months ahead.

“This is a very tough business,” agrees CPC's head of forecast operations, climatologist Edward O'Lenic. But he says Livezey and Timofeyeva's analysis of past skill “does have some flaws” that make it underrate CPC's performance, and he thinks some of the forecast tools they dismiss may still prove useful in ways researchers don't yet understand.

Climatologist Anthony Barnston of Columbia University's International Research Institute for Climate and Society in Palisades, New York, leans toward what he calls O'Lenic's “philosophical” preference for being more inclusive of forecasting tools. But Barnston agrees with Livezey that modeling holds the greatest promise for improving seasonal forecasting.

3. APPLIED PHYSICS

# Bizarre 'Metamaterials' for Visible Light in Sight?

When, in 2000, physicists unveiled the first “left-handed metamaterial”—an assemblage of metallic rods and rings that interacted with and bent microwaves in strange ways—physicists immediately knew they had a grand goal to shoot for: miniaturized metamaterials that would bend visible light in the same way. If such things could be made, they could result in wild devices, such as a “superlens” that would focus light tighter than any conventional lens. Metamaterials might be used to make invisibility cloaks, too, researchers have since shown. Now, metamaterials for visible light may be within reach, thanks to advances reported this week online in Nature and on page 930 of this issue of Science.

Both results come from the lab of Xiang Zhang, an applied physicist at the University of California, Berkeley. In Nature, Zhang's team describes a metamaterial that works for near-infrared light and, unlike previous materials, is three-dimensional. In Science, the team presents a different three-dimensional metamaterial that bends visible red light in the desired way.

Opinions vary as to how substantial the advances are. “With the Science paper, we are really very, very close” to applications with visible light, says Costas Soukoulis, a physicist at Iowa State University in Ames and the Department of Energy's Ames Laboratory. But Henri Lezec, an electrical engineer at the National Institute of Standards and Technology (NIST) in Gaithersburg, Maryland, says “the claims are misstated and overhyped.”

Metamaterials put a kink in the way light usually passes from one medium into another. Suppose light from the setting sun shines on a pond. As light waves strike the surface, their direction will change so that they flow more directly down into the water. (See diagram.) Such “refraction” arises because the light travels more slowly in water than in air, giving water a higher “index of refraction.” Still, the light continues to flow from west to east. Were water a left-handed metamaterial, however, “negative refractions” would bend the light back toward the west.

To produce the effect for near-infrared light, Zhang, Jason Valentine, and colleagues created a material that looks like a miniature waffle. They laid down 21 alternating layers of conducting silver and insulating magnesium fluoride on a quartz substrate and drilled holes in the stack using an ion beam. They cut the stack at an angle to make a prism and showed that it bent light the “wrong” way compared with an ordinary prism. To achieve negative refraction in the visible range, Zhang, Jie Yao, and the team used a standard electrochemical technique to make a sample of aluminum oxide filled with a regular array of nanometer-sized holes, which they filled with silver. When they shined red light onto the sample at an angle, it underwent negative refraction.

That might seem to seal the deal, but not everyone is convinced. Lezec argues that the infrared metamaterial isn't truly three-dimensional because it works for light coming from only a narrow range of directions. The metamaterial that bends visible light works for light of only a single polarization, he notes. And all agree that, strictly speaking, it does not have a key property—a negative index of refraction—although the infrared metamaterial does.

That's nitpicking, says Vladimir Shalaev, a physicist at Purdue University in West Lafayette, Indiana. “What's wrong with [using] a particular polarization?” he says. “As a first step, it's not so bad.” The real advance in the Science paper may be a new self-organizing approach to fashioning the materials, Shalaev says. Soukoulis warns that researchers must confront a basic problem: At shorter wavelengths, metamaterials absorb far too much light. For now, however, the future for metamaterials looks particularly bright.

4. # HIV/AIDS: Treatment and Prevention Exchange Vows at International Conference

1. Jon Cohen

MEXICO CITY—AIDS researchers have long argued that HIV prevention and treatment efforts should go hand in hand, but they rarely do. Their fickle relationship received intense scrutiny at the XVII International AIDS Conference held here last week. “They keep going to the altar,” said Myron “Mike” Cohen of the University of North Carolina (UNC), Chapel Hill, in a plenary presentation. “They never get married. They have to get married today.”

More than 20,000 researchers, health care workers, representatives from hard-hit communities, and activists attended the conference, which had never been held in Latin America before. The meeting ran 3 to 8 August, and about one-fourth of the participants came from the region.

As usual at these gatherings, science shared the limelight with diverse issues such as scaling up access to anti-HIV drugs, the increasing criminal prosecution of people who infect others, and the need for countries to address their epidemics in ostracized groups. Protests were more muted than in past years, although several added a novel Latin American spice to this conference staple.

New research findings were fewer and farther between than ever, creating the sense that the meeting has evolved into a giant review paper rather than a place for colleagues to share their latest data. “This is more a world AIDS summit, where every 2 years we reexamine everything we know,” said Julio Montaner, the new president of the International AIDS Society (IAS), the meeting's organizer.

Cohen was one of several presenters who stressed that the great gains in treatment have overshadowed prevention needs. Today, 3 million people in low- and middle-income countries receive anti-HIV drugs, but an estimated five people become infected for every two on treatment. “There has not been that push for prevention as there's been for treatment,” said Peter Piot, head of the Joint United Nations Programme on HIV/AIDS. “If we thought the first phase was hard, we have to prepare for even tougher times.”

Piot also noted that the characteristics of the epidemic keep changing in different locales, urging countries to “know their epidemics” and target prevention to the most vulnerable groups. In Thailand, where the epidemic has been concentrated among injecting drug users and sex workers, married women now account for more new infections than any other group. In parts of sub-Saharan Africa, where epidemics have been primarily driven by heterosexual sex, injecting drug use is an increasingly important mode of spread. China, which has a large number of infected injecting drug users, today has a growing epidemic in men who have sex with men. In the United States, infections of whites peaked in the mid-1980s; blacks now account for 45% of the new infections and have an eight times higher risk of becoming infected, according to new estimates published by the U.S. Centers for Disease Control and Prevention (CDC). “The end of AIDS is nowhere in sight,” said Piot.

View this table:

The success with combinations of potent anti-HIV drugs, which reduce the amount of virus people carry and make them less infectious, has led to the increasing awareness that treatment is prevention, both for individuals and populations. But the degree to which the drugs can prevent infections has proved highly contentious.

A statement issued by the Swiss Federal Commission for HIV/AIDS in January on this topic served as a lightning rod. After reviewing the scientific literature, the Swiss commission concluded that a heterosexual person faced virtually no risk of becoming infected by having unprotected sex with an HIV-infected person on continued treatment, provided that person had undetectable levels of virus in the blood for 6 months and no sexually transmitted infections. The statement stopped short of explicitly discounting the value of condoms, but many thought that was its implicit message.

“There's condom absolutism, and everyone who questions it is put into controversy,” said Bernard Hirschel, who heads the HIV/AIDS program at the University Hospital, Geneva. The main aims of the statement, he said, were to tell “discordant” couples—in which one is infected and the other isn't—who met these criteria that they could safely try to have children and also to combat a Swiss law that says an HIV-infected person who has sex without a condom can be held criminally liable, even in the absence of infecting a consenting partner.

Kevin De Cock, head of HIV/AIDS for the World Health Organization, and others blasted the statement as irresponsible. “It just doesn't seem like a cautious public health recommendation,” said De Cock. “I don't think anyone's shown the threshold below which people cannot transmit.”

A model published in the 26 July issue of The Lancet by David Wilson and colleagues at the University of New South Wales in Sydney, Australia, further emphasized the dangers. The study devised a mathematical model to compare 10,000 discordant couples that had unprotected sex for 10 years with the same number of couples who used condoms 80% of the time. The risk of transmission increased four times in the unprotected group because of occasional viral rebounds that happen to people on effective treatment.

Also hotly contested was the degree to which ongoing treatment can prevent transmission on the population scale. IAS President Montaner, a researcher at the University of British Columbia, co-authored an article in the 1 July issue of the Canadian Medical Association Journal that contends that potent treatment led to a decrease in HIV's spread in British Columbia. Specifically, their study notes that new HIV infections dropped about 50% in British Columbia from 1995 to 1998, the years when highly potent anti-HIV drugs first became available. During the same years, syphilis infections increased, suggesting that the drop was not due to condom use or other behavioral changes. “Antiviral therapy greatly lowers infectiousness,” contended Montaner.

But epidemiologist Geoffrey Garnett of Imperial College London countered that antiretroviral drugs are unlikely to make a large impact on transmission on a global scale. Roughly 80% of infected people do not even know their status. Of those who do, most are not eligible for free treatment until their immune systems have been substantially damaged—which means most transmissions occur long before people are taking the drugs.

Garnett and others encouraged their colleagues to embrace the notion of “combination prevention.” No currently available intervention can by itself turn an epidemic around, but by combining treatment with preventive measures such as condoms and circumcision, it may be possible to create “a natural synergy,” Garnett said. “Rather than arguing for a single magic bullet, we really need to be trying to focus everything that we can on what works to realize these natural synergies.”

The growing enthusiasm for combination prevention in part reflects the dispiriting fact that the vast majority of biomedical prevention studies, from large human vaccine trials to microbicides to treatment of sexually transmitted diseases, have failed (see table). Still, many investigators have high hopes for what could be something of a magic bullet: pre-exposure prophylaxis (PrEP), which gives anti-HIV drugs to uninfected people. The idea is that people at high risk of infection will take the drugs shortly before having sex, much in the way that people take antimalarial drugs before visiting countries where that disease is prevalent. Studies around the world are now enrolling more than 18,000 people to test this concept—more than the number of people in AIDS vaccine trials, noted Mitchell Warren, head of the AIDS Vaccine Advocacy Coalition in New York City. UNC's Cohen predicted that PrEP, similar to the successful strategy used to prevent transmission of HIV from an infected, pregnant woman to her baby, “is almost certain to work.”

The approach has had remarkable success in monkeys. Walid Heneine of CDC in Atlanta, Georgia, described experiments in which he and his colleagues inserted anti-HIV drugs into the vaginas of six monkeys and then 30 minutes later tried to infect the animals with vaginal infusions of an engineered AIDS virus. None of the animals became infected after 20 such “challenges,” whereas seven of eight untreated control animals did.

Anthony Fauci, head of the U.S. National Institute of Allergy and Infectious Diseases in Bethesda, Maryland, said PrEP may lead to protection in more ways than one: The drugs prevent infections by killing or weakening the AIDS virus, which could trigger immune responses that subsequently derail infections. “That may be the first vaccine,” said Fauci.

5. ARCHAEOLOGY

# Going Deeper Into the Grotte Chauvet

1. Michael Balter

Ten years of research have yielded detailed new insights into the stunning images considered the world's oldest cave art. But questions about their age are resurfacing.

Ten years of research have yielded detailed new insights into the stunning images considered the world's oldest cave art. But questions about their age are resurfacing

Sometime during the last ice age, artists entered a cave in southern France, lit torches and fires, and began work on a masterpiece. Squatting on the cave floor and wielding pieces of charcoal, the artists first drew the outlines of two rhinoceroses locking horns. Then, standing up and moving to the left, they sketched the heads and upper bodies of three wild cattle. Finally, a lone artist stepped forward to execute the pièce de résistance: four horses' heads, drawn with exquisite shading and perspective in the center of the tableau, each horse displaying its own expression and personality.

This, at least, is how researchers studying the Chauvet Cave in the Ardèche region of southern France envision the creation of the famous Horse Panel. According to direct radiocarbon dating of the two rhinos and one of the cattle, they were drawn between 32,000 and 30,000 radiocarbon years ago, making them the oldest known cave art in the world. (The exact calendar age is uncertain because there is no accepted radiocarbon calibration for this period; see Science, 15 September 2006, p. 1560.) These early dates, announced soon after the cave's discovery in December 1994, struck a blow to conventional assumptions that such sophisticated artworks did not appear until up to 15,000 years later.

In the decade since researchers began working in the Grotte Chauvet (Science, 12 February 1999, p. 920), they have photographed and redrawn many of the more than 400 animals depicted, identified signs of human activity such as footprints and hearths, deciphered the cave's geology, and analyzed thousands of bones left by cave bears that shared the cave with humans. And archaeologists have begun to propose hypotheses about what the art might have symbolized to those who created it.

But as the team continues its work, a small but persistent group of archaeologists continues to question the age of the paintings. “Chauvet is the world's most problematically dated cave art site,” says archaeologist Paul Pettitt of the University of Sheffield, U.K., whose most recent challenge was published online this month in the Journal of Human Evolution (JHE). That contention—which the team vigorously rejects—has critical implications for our understanding of the origins of art. “The fundamental importance of Chauvet is to show that the capacity of Homo sapiens to engage in artistic expression did not go through a linear evolution over many thousands of years,” says cave art expert Gilles Tosello of the University of Toulouse (UT), France. “It was there from the beginning.”

## Lions, and horses, and bears, oh my!

Since resolving lawsuits and beginning scientific study a decade ago, researchers have reconstructed how the artists worked, analyzing each stroke of charcoal, red ochre, and engraving. Tosello and his wife, UT cave art expert Carole Fritz, have spent hundreds of hours perched in front of the 6-square-meter Horse Panel, photographing it in sections and drawing the artworks onto tracing paper. Working in this meticulous fashion, and noting the superposition of charcoal lines as well as slight thickenings at the beginning and end of each stroke, the pair was able to reconstruct the order and direction in which each line was drawn.

“The detailed nature of their observations is extraordinary,” says archaeologist Iain Davidson of the University of New England in Armidale, Australia. Tosello and Fritz found that the artists who drew the two rhinos began with the horns and muzzles, then drew the front legs and bellies, and finally the rest of the bodies, making corrections and filling in details as they went. As the artists worked around the panel from the edges to the middle (see diagram above), they reserved a space in the center for the four horses, whose heads and necks are slightly superimposed over the backs of the cattle and arranged in a tight, diagonal orientation. This suggests to Tosello and Fritz that they were drawn by one artist. To make the horses' heads even more vivid, the artist used a tool to etch the cave wall around their muzzles so that they stand out in a prehistoric version of bas-relief.

“The entire composition is very homogeneous and has a very strong coherence,” Tosello says, making it likely that the artwork was drawn by a small number of artists over a fairly short period of time. He adds that the Horse Panel, along with other compositions in the cave—such as a troop of lions apparently chasing a herd of bison—seems to be telling a story. “The animals appear on the wall in a certain order, like characters coming on stage during a play,” he says. He speculates that prehistoric humans, who hunted bison, might have identified with the lions and wished to emulate their hunting prowess.

Humans probably kept their distance from lions, but the artists of Chauvet shared their cave with at least one dangerous animal: the cave bear. The team has found about 4000 cave bear bones, representing nearly 200 animals, on the cave floor, including a skull that was apparently placed deliberately atop a limestone block. Archaeologists have long debated whether humans hunted cave bears, worshipped them, or had some other relationship with these now-extinct animals. The artists clearly saw them from time to time: Chauvet's menagerie includes 15 drawings of cave bears.

Radiocarbon dates on 18 bear bones put them between 28,850 and 30,700 radiocarbon years ago, “slightly younger” than the dates for the paintings, according to evolutionary biologist Hervé Bocherens of the University of Tübingen in Germany. One other bone exposed by erosion of the cave floor was dated to 37,000 years ago, indicating, Bocherens's team concluded in a 2006 paper in JHE, that bears were already using the cave when prehistoric artists first entered.

“Imagine the terror of entering the cave with flickering lights, knowing that there might be bears in there,” says Davidson. But bears and humans might have visited the caves in different seasons—winter hibernation for the bears, spring for the humans, points out paleogeneticist Jean-Marc Elalouf of the French Atomic Energy Commission in Saclay.

## How old is old?

The dates for both the bears and the art correspond to the Aurignacian period, the first culture associated with the modern humans who colonized Europe beginning about 40,000 years ago. Yet some researchers have argued that the art more closely resembles much later cultures, possibly even the Magdalenian, which stretched from about 17,000 to 12,000 years ago and to which the great paintings at Lascaux in France and Altamira in Spain are attributed. But most experts accepted the dates, which were produced by the Laboratory of the Sciences of Climate and the Environment (LSCE) in Gif-sur-Yvette, France, a lab that pioneered the direct dating of cave paintings.

In 2003, however, Sheffield's Pettitt, along with British archaeology writer Paul Bahn, threw down the gauntlet again, arguing in Antiquity that the dates were not reliable because they had not been replicated by other labs; the Chauvet team defended its results in the same issue. “Chauvet is the best dated rock art site in the world,” says French rock art expert Jean Clottes, former leader of the Chauvet team. Randall White, an archaeologist at New York University, agrees: “There are more dates from Chauvet than from most other caves combined.”

In his new JHE paper, Pettitt launches the most detailed onslaught yet, saying that the drawings are simply too magnificent for that time. “Chauvet stands out in terms of overall technical sophistication whatever one compares it to,” Pettitt told Science. He insists that the seven direct dates from paintings are unreliable because of the small sample sizes and the possibility of contamination from the cave wall.

Pettitt also discounts radiocarbon dates from more than 40 charcoal samples from the cave floor, which range between about 27,000 and 32,000 years ago, as well as recent redating of charcoal samples from a chamber rich with art. Those samples, split between six radiocarbon labs, gave consistent results of about 32,000 years before the present. Pettitt says these charcoal dates are irrelevant to the age of the art. “Could I not enter the cave today, pick out a piece of this well-preserved charcoal from a hearth on the floor, and write ‘Paul Pettitt was here'on the cave wall?”

Some archaeologists take Pettitt's argument seriously. “People might have picked up old charcoal from the Aurignacian period during the Magdalenian,” says William Davies, an Aurignacian expert at the University of Southampton, U.K. Pettitt's Sheffield colleague, archaeologist Robin Dennell, goes further: “Chauvet should be removed from assessments of early modern humans in Europe. Including it leads to a gross distortion of their cognitive abilities.”

But the Chauvet team is having none of it. “This is ridiculous,” Clottes says. “There were heaps of charcoal right in front of the paintings.” Tosello agrees: “Who can believe that the Aurignacians came into the cave, left behind piles of charcoal without making any drawings, and then thousands of years later the Magdalenians entered and used the charcoal kindly left by their ancestors to draw on the walls?” Team members insist that the close agreement of dates from the paintings, the charcoal, and the bear bones argues that the cave was frequented by humans and bears during the Aurignacian, not the Magdalenian. Clottes also cites uranium/ thorium dating that suggests that the cave entrance was blocked to entry by a landslide about 19,000 years ago—before the Magdalenian period. As for replicating the direct dating of the paintings, Hélène Valladas, leader of the LSCE team that carried out this work, says it is not possible to take more samples without “visibly altering the [art] traces.”

Some archaeologists also find Pettitt's stylistic arguments unpersuasive. Even Davies, who hesitates to call the art Aurignacian, says, “I am not convinced the paintings are Magdalenian. … Some of the techniques are unique to the site and not found in the Magdalenian period.” White adds that there is plenty of other evidence for sophisticated symbolism in the Aurignacian, including thousands of personal ornaments made from shell and bone. “It's all part of the Aurignacian package,” White says.

In any case, the significance of Chauvet goes beyond the “oldest art” debate, says anthropologist Margaret Conkey of the University of California, Berkeley. “Chauvet was an expression of the sensibilities, beliefs, and social relations of anatomically modern humans in this part of the world,” she says. “What was it about their lives that made imagemaking in caves meaningful?”

6. OLIVERA FINN PROFILE

# Directing a Life in Science

1. Mitch Leslie

After forgoing theater ambitions, and despite early marriage and motherhood, Olivera Finn has risen through immunology's ranks thanks to her work on cancer vaccines.

After forgoing theater ambitions, and despite early marriage and motherhood, Olivera Finn has risen through immunology's ranks thanks to her work on cancer vaccines

Take a look at Olivera “Olja” Finn's life, and you can tick off the actions women are supposed to avoid if they want to advance in science. Get married fresh out of high school. Check. Interrupt your education for your husband's sake. Check. Allow his career to take precedence over yours. Check. Have children before you have a job and give birth at what seem like inopportune times, such as shortly before you start graduate school. Check.

Yet Finn has, with great success, pursued career and family goals simultaneously. She celebrated her 40th wedding anniversary last month, has raised a daughter and a son, and, at the age of 59, already has grandchildren. Professionally, Finn has prospered. Nearly 20 years ago, she discovered the first cancer antigen, a tumor molecule that elicits a reaction from immune cells. And despite spending her youth in Communist-run Yugoslavia, Finn has climbed the academic ladder in the United States—she is chair of immunology at the University of Pittsburgh in Pennsylvania and has served as president of the American Association of Immunologists. She argues that interweaving career and family is essential. “I don't think we live long enough to do things sequentially.”

Colleagues laud her work in cancer immunotherapy, the goal of which is to enlist the immune system to combat tumors. In an extension of her tumor antigen discovery, Finn's group is gearing up to test a vaccine to prevent benign colon growths from spawning deadly cancers. Her effort is rare in that most cancer “vaccines” are not preventive; they're designed to treat serious tumors. The few preventive cancer vaccines approved for use target tumor-causing pathogens such as the hepatitis B and human papilloma viruses rather than growths themselves, as Finn's vaccine does.

“The field has advanced faster because of her,” says Martin Cheever, a medical oncologist at the Fred Hutchinson Cancer Research Center in Seattle, Washington. Finn deserves credit not only for her scientific insights, he adds, but also for her devotion to nurturing other scientists' research and fostering cross-disciplinary collaborations. Without such prompting, “cancer biologists and immunologists [usually] sit on their own sides of the fence,” notes immunologist Ralph Steinman of Rockefeller University in New York City.

The courage, tenacity, and independent-mindedness Finn needed to start anew in a strange country also characterize her science, says Paola Castagnoli, scientific director of the Singapore Immunology Network and Finn's friend since the late 1970s. Finn is currently exploring the provocative idea that infections throughout life, including chickenpox and other childhood diseases, prime our defenses against cancer. “She is a very good scientist because scientists should not be conformists,” says Castagnoli.

## The accidental scientist

Growing up in what was then Yugoslavia, Finn aspired to direct plays. But she strayed from the script once she met Seth Finn, an American college student on a foreign exchange program. Over her parents' objections, the couple married and moved to the United States. She'd been studying English since age 7, so language wasn't a barrier. What shocked her, she says, was Americans' ignorance of foreign affairs, obsession with money, and willingness to make long-haul commutes.

After briefly attending college in California and Indiana, she ended up in Puerto Rico, where her husband was serving in the Coast Guard. At the urging of her father, a theater manager with geology and biology degrees, Finn had followed the technical track at her Yugoslavian high school. In Puerto Rico, her scientific ambition blossomed. For an undergraduate project at the Interamerican University in San Juan, where she completed her bachelor's degree in biology, Finn figured out missing steps in the life cycle of a hookworm that circulates among humans, birds, rats, and cockroaches. The work involved poking around seedy areas of downtown San Juan and picking up roaches as big as a tablespoon, but she loved it. “The life of research—getting data and making hypotheses—consumed me,” she says.

After finishing a Ph.D. and a postdoc at Stanford University in Palo Alto, California, Finn set up her own lab at Duke University in Durham, North Carolina. She chose Duke because Seth, who by that point had earned a Ph.D. in communications from Stanford, had landed a position at the nearby University of North Carolina, Chapel Hill. When she arrived at Duke in 1982, it was a hotbed of transplant immunology research, and she focused on identifying what triggers the rejection of donated organs. Her group reared T cells extracted from patients who'd received kidney transplants and nailed down which of the donor's antigens, or molecular markers, provoked the cells to attack. Although hundreds of molecules could potentially prompt a rejection response, typically only one or two antigens did, her team discovered.

That success spurred Finn to ask whether the same techniques might shed light on cancer-immune system interactions. Scientists had known since the 1950s that cancer cells can rouse the immune system. In fact, a debate has raged since then about whether the immune system thwarts many incipient cancers, or whether the immune response is too feeble to curb most abnormal growths. However, in the early days of this debate, scientists didn't even know what antigens on tumors trigger an alarm.

In the mid-1980s, Finn decided to track down these telltale tags. Looking back, the decision to shift to tumor immunology was naïve, Finn says. The lab's skill in identifying rejection antigens “gave us a confidence that was exaggerated.” Finding cancer antigens turned out to be much tougher. For one thing, whereas a transplanted organ riles the immune system, tumor cells elicit a much weaker response.

Weak, yes, but not undetectable, and by 1989 Finn's lab had nabbed the first cancer antigen, a protein called MUC1 that protrudes from pancreatic and breast tumor cells. Human T cells keyed on this antigen, her team reported.

MUC1 also decorates normal cells in several organs, so why don't T cells pounce on those tissues? The answer came in work Finn continued after moving to the University of Pittsburgh in 1991. Normal MUC1 is festooned with carbohydrate chains, which are nearly absent from the protein fashioned by cancer cells. The pattern is clear, Finn says. Tumor antigens usually differ from their normal counterparts in some way, such as structure, quantity, or cellular location. For cyclin B1, which helps propel cells through mitosis, quantity explains why it can act as a tumor antigen. In normal cells, the amount of cyclin B1 remains low except for a spike at the beginning of mitosis. Yet cancer cells churn out the protein nonstop. MUC1, cyclin B1, and the like are not “self” antigens but “abnormal self” antigens, Finn says.

## Family time

As Finn talks about her life, you don't hear any regrets—she clearly doesn't regard her early marriage and motherhood as youthful indiscretions. Finn, who started graduate school at Stanford with a 7-month-old son to tend, encourages women at the same stage of their careers to have children. If you think you'll have more time for parenting later in life, you are wrong, she says.

Carrie Miceli, who was Finn's first graduate student and is now an immunologist at the University of California, Los Angeles, says she followed Finn's example, although she waited until starting her own lab to have a child. “It was great to see a woman with kids and a family who was not talking about what a compromise it was,” says Miceli.

Finn and her husband took turns going for advancement. After Seth's job led them to North Carolina, the choice to move to Pittsburgh was hers. For 4 years, Seth commuted every week between Pennsylvania and North Carolina before being hired by Robert Morris University in Pittsburgh.

## The cancer shot

For more than a decade, Finn has worked to package the tumor antigen she discovered into a vaccine that would prevent cancer. Her group conducted initial safety trials of a MUC1-containing vaccine, using patients with advanced pancreatic cancer. In 2005, for instance, the researchers reported that the vaccine produced no obvious side effects—and also seemed to promote an immune response to MUC1 in some recipients. Yet the U.S. Food and Drug Administration (FDA) balked at her proposal to test the vaccine in healthy people, she says, partly because of the fear that it would trigger autoimmunity, an immune assault on normal tissue.

Now she's finally getting a chance. This summer, her group will launch a 5-year trial to determine whether injections containing abnormal MUC1 can prevent recurrence of intestinal adenomas. Surgeons usually remove these benign growths because they can morph into colon tumors. However, adenomas often sprout again after the operation. The study's control group will be historical: past patients who were operated on by the same doctors. Finn concedes that even this trial isn't ideal. The researchers are testing the vaccine's ability to prevent adenoma regrowth, not its ability to fend off cancer in healthy people. Moreover, the patients will be elderly, and the response to vaccines dwindles with age.

Age's affect on immunity also figures into an idea that has captured Finn's interest. Work by her group and other labs suggests that many of us receive “natural” vaccinations against cancer from an unexpected source: pathogens. A variety of body invaders, including those that cause childhood diseases, spur the production of the same abnormal self antigens as cancer cells. The chickenpox virus, for instance, sparks an explosion in cyclin B1. The mumps virus prompts cells to display denuded MUC1. Getting sick in our youth, when our immune systems are primed to make the memory cells that can confer lifelong immunity, might spare us from cancer later on, Finn proposes.

To test the idea, Finn teamed up with epidemiologist Daniel Cramer of Brigham and Women's Hospital in Boston and colleagues. They found that women who'd undergone events that can lead to infections or inflammation—including intrauterine device use, pelvic surgery, and broken bones—were more likely to carry antibodies to MUC1, a sign of an immune response. These women also had a lower risk of developing ovarian cancer, the researchers reported in 2005.

In a life full of challenging career moves, Finn is pondering her next and last. She says she would like to work at FDA to help pave the way for preventive cancer vaccines. “People used to say it would take 10 years to evaluate [these] vaccines, but it's been 10 years and we are still discussing how it will take 10 years.” As in her family life, Finn is not inclined to wait.

7. U.S. HIGHER EDUCATION

# Science Scholarships Go Begging

1. Jeffrey Mervis*
1. With reporting by Fayana Richards.

Despite ever-rising college costs, a $4.5 billion federal aid program to lure students into science is vastly undersubscribed. Despite ever-rising college costs, a$4.5 billion federal aid program to lure students into science is vastly undersubscribed

Paul McCarl dropped out of Brigham Young University (BYU) in 1991 and rode the dotcom boom and bust writing entertainment software before moving into retail management. But when his son told him that his eighth-grade teacher had said the phases of the moon are caused by Earth's shadow falling on the lunar surface, McCarl decided he was needed in the classroom. So 2 years ago, at the age of 38, the former computer science major returned to BYU's Provo campus and enrolled in its physical science teacher program. And this week, he began his new career in the science department at Whitehorse High School, a tiny school on the Navajo reservation in southeastern Utah.

McCarl would seem like the perfect candidate for a fledgling federal scholarship designed to attract more U.S. students into scientific fields. But he was excluded because the BYU courses he was taking didn't meet its stringent eligibility requirements. In fact, the bar is so high that the Department of Education is spending money at only half the rate Congress envisioned in 2006 when it created the 5-year, $4.5 billion National Science and Mathematics Access to Retain Talent (SMART) and the Academic Competitiveness (AC) grant programs. Secretary of Education Margaret Spellings says the reason so few college students are eligible for the largest federal aid program of its kind is that they haven't taken the necessary courses in high school. But university financial aid directors point to the many requirements, a break from the traditional practice of awarding aid according to financial need. “The AC and SMART grants are the most administratively burdensome programs that I have ever seen,” says Katy Maloney, director of financial aid at the University of California, Davis. “It's pretty much a nightmare because they have so many rules.” The grants were created in response to a flood of reports on the woeful condition of U.S. math and science education and the need for a more technically trained workforce. To be eligible for the AC grant, worth$750 in the first year and $1300 in the second year, students must qualify for the government's major needs-based scholarship, called a Pell grant, and have graduated from “a rigorous secondary school program.” That means 3 years each of higher level math and science and at least 1 year of a foreign language. Once in college, they also need to maintain a 3.0 or higher grade point average. The SMART grant pays$4000 a year to third- and fourth-year students with good grades who are pursuing majors in the sciences, mathematics, engineering, and technology.

With seven children and a wife to support, McCarl was counting on SMART grants to complete his college education. But he got tripped up by the provision that a student's major must be on a list approved by the department. Although his degree will be awarded by the College of Physical and Mathematical Sciences, whose programs are eligible, his course of study falls under the category of secondary education, which doesn't qualify.

Such requirements are one reason why, despite the ever-rising cost of college, the money for AC and SMART grants isn't flying out the door. The department spent barely half of its $850 million allocation in 2006–07, awarding grants to 360,000 students. That shortfall caused Congress to cut the 2007–08 allocation to$397 million. It's also elicited a promise from Spellings to double the number of grant recipients by its final year (2010–11).

That's 2 years after she and her boss, President George W. Bush, leave office, of course. In the meantime, Spellings blames the underutilization on the sorry state of elementary and secondary education and argues that the best way to raise participation rates is to reauthorize the president's signature education initiative, No Child Left Behind.

But financial aid directors question whether the prospect of a small scholarship is likely to induce students to take more math and science courses before they enter college. “Let's be realistic,” says Anna Griswold, who oversees student aid programs at Pennsylvania State University in State College. “Is a high school sophomore going to take a tougher schedule because he might get \$750 more as a college freshman?” She and other student aid officials agree that the SMART grant might be more of a lure for some upper-level students, but they say its impact would be very difficult to measure.

The chair of the House Committee on Education and Labor, Representative George Miller (D-A), declined to speculate on the fate of the scholarship program in the next Congress. “Let's give it some time and see what happens,” Miller said last month after chairing a hearing on corporate efforts to improve STEM education. A spokesperson for the Republican minority on the committee predicted that legislators won't take a hard look at funding levels for the programs until it's time to refill the pot.

Even so, Congress this year passed two bills that are expected to goose participation rates. In May, it decided that half-time students and permanent residents were eligible for both programs. And last month, in a long-overdue higher education bill awaiting the president's signature, it gave state education officials the authority to certify a rigorous course of study, a power that previously had rested with the education secretary.

Neither will affect McCarl, who this month moved his entire family to the reservation. But he's okay with that. “I'm getting the chance to become a teacher,” he says. “And I plan to stay here for the rest of my life.”

8. GLOBAL WARMING

# Climate Change Hot Spots Mapped Across the United States

1. Richard A. Kerr

Taking some of the fuzziness out of climate models is revealing the uneven U.S. impact of future global warming; the most severely affected region may be emerging already.

Taking some of the fuzziness out of climate models is revealing the uneven U.S. impact of future global warming; the most severely affected region may be emerging already

Now that almost everyone expects a certain amount of global warming by the end of the century, attention can turn to more local climate change. What's going to happen in our own backyards? Researchers can't go that far yet, but in an effort to squeeze the maximum detail out of notoriously fuzzy climate models, they are pooling results from some of the most sophisticated simulations available.

The latest regional climate effort points up the uneven burden climate change will place on the United States. “It highlights that there are regions where climate changes will be bigger than others,” says climate modeler Gerald Meehl of the National Center for Atmospheric Research (NCAR) in Boulder, Colorado. The American Southwest looks to be hardest hit by far, but the work also highlights a dramatic increase in year-to-year climate variability contributing to hot spots.

The new work is in press in Geophysical Research Letters (GRL). As climate modeler Noah Diffenbaugh of Purdue University in West Lafayette, Indiana, and his colleagues lay out in the paper, regional climate modeling in the wake of last year's report from the Intergovernmental Panel on Climate Change has come a long way since the previous IPCC report in 2001. For that report, researchers divided the contiguous 48 states into 1300-kilometer-wide west, central, and east regions, including a good bit of Canada in the west. Drawing on IPCC simulations of future greenhouse climate generated by nine then-state-of-the-art global climate models, they concluded that each broad region could expect slightly more warming and in the winter slightly more precipitation than the global average.

In the GRL paper, Diffenbaugh and his colleagues offer a much sharper picture of climate change. They combine forecasts from 15 new, state-of-the-art global models run for last year's IPCC report. These models individually paint a more detailed picture than their predecessors did and have more realistic renditions of the physical processes in the climate system. The group also formulates a new gauge of climate change—climate responsiveness—by combining projected changes in temperature and precipitation as well as changes in variability of those climate properties from year to year. High values of this climate responsiveness mark “hot spots” where the models say climate will be changing the most.

According to the 15-model consensus, the strongest U.S. hot spot by far stretches across the Southwest from southern California to west Texas and intensifies even more over northern Mexico. By another statistical analysis technique, the American Southwest hot spot extends northward into Nevada, Utah, and Colorado. By either technique, the U.S. Southeast is a distinct “cool spot,” a region relatively less responsive in changing temperature and precipitation, although Diffenbaugh cautions that “we need to be careful to not overinterpret these areas as ‘safe’ or ‘immune.’” Other studies have suggested that these less responsive regions may be at risk of other sorts of greenhouse changes, such as increased severe weather in the Southeast.

Two higher resolution models not included in the consensus—one global, the other an extremely high-resolution model of the continental United States—suggest a similar pattern but also identify a milder climate change hot spot in the Midwest.

Most surprising to Diffenbaugh, the better part of a hot spot's strength came not from progressive warming or a long-term rise or fall in precipitation but from increased variability from one year to the next, especially in precipitation. Models have predicted that a strengthening greenhouse would make the climate more variable, but “I'm not sure what that means,” says regional climate modeler Linda Mearns of NCAR. “More attention should be given to how variability is going to change.”

“Needless to say, this work is only the beginning of a possible new avenue … towards a clearer picture of where regional climate change matters,” regional modeler Jens Christensen of the Danish Meteorological Institute in Copenhagen writes in an e-mail. It was good that the group checked the combined global models against the higher resolution models, he explains. But the work points up the need for combining results from multiple regional models, not just the global models. Such an approach might help address concerns that the models still aren't very good at replicating climate change across the United States during the past 50 years, as meteorologist Kevin Trenberth of NCAR notes in an e-mail. That may be in part because the models have trouble simulating natural climate changes induced by slow changes like El Niño, he says.

Shortcomings or not, the IPCC models may have found a hot spot that is already developing. The predicted Southwest hot spot of climatic change looks much the same during the next 30 years as at the end of this century. And that future hot spot bears a strong resemblance to the drying and warming of the Southwest during the past decade or so. Says Diffenbaugh: “We may already be seeing some emerging hot spot patterns.”