# News this Week

Science  27 Jun 2008:
Vol. 320, Issue 5884, pp. 200
1. ARCHAEOLOGY

# Early Stonehenge Pilgrims Came From Afar, With Cattle in Tow

1. Michael Balter

Each year, nearly a million tourists are drawn to Stonehenge's massive stone pillars. Now evidence fresh from the lab suggests that the iconic monument in southern England was a place of pilgrimage even in prehistoric times. But why they came remains a mystery.

Isotopic studies of teeth from six cattle found at a nearby earthen henge show that the animals were herded to the site from distant parts of Britain. Some of the animals apparently came from as far away as Wales—also the origin of Stonehenge's smaller “bluestones.” The unpublished results, coming on the heels of new dates for human remains at Stonehenge, are fueling ongoing debates about whether the 5000-year-old monument served chiefly as a “place of the dead” or whether its stones were valued for their healing properties.

The cattle findings “are potentially exciting,” says Mike Pitts, editor of British Archaeology, because they “imply an unusually large social and political network behind [Stonehenge's] creation.” Archaeologist Alasdair Whittle of Cardiff University in the U.K. says the new data support evidence that Stonehenge was preceded by “centuries of movement and connections” among prehistoric peoples in Britain. But both researchers note that the results are still preliminary. “Six [animals] is a small sample, so caution is appropriate,” says Whittle.

The cattle teeth come from ancient rubbish deposits at a site 3 kilometers away called Durrington Walls. This massive circular earthwork, about 500 meters in diameter, has long been thought to be related to Stonehenge, which itself was first built as an earthwork.

The teeth were analyzed by Sarah Viner, a graduate student with University of Sheffield zooarchaeologist Umberto Albarella. Viner, in collaboration with Jane Evans of the Natural Environment Research Council's Isotope Geosciences Laboratory in Nottingham, U.K., looked at the ratio of two strontium isotopes, 87Sr and 86Sr, in the teeth of adult animals. This ratio varies depending on what type of soil the animals have grazed on: Higher ratios stem from Britain's older geological formations, and lower ratios arise in the younger chalklands of southern England where Stonehenge is located (see map). The strontium “signature” is laid down when the teeth are formed in growing animals and does not change later.

Viner and Evans found that none of the six animals was raised in the chalklands. Two appeared to come from Wales or Scotland, the others from locations in England that could not be determined.

Sheffield archaeologist Michael Parker Pearson, who led excavations at Durrington Walls, says these results fit other signs that people came to the site only periodically. “There is no evidence of cereals or grinding stones,” he says, as would be expected at a site permanently occupied by the early farmers who then inhabited Britain. Moreover, Albarella says, none of the cattle or pig bones were from very young animals, making it unlikely that animals were raised at the site. Instead, he believes they must have been brought there from afar, perhaps during episodes of feasting at the summer and winter solstices.

Parker Pearson says the findings support his hypothesis that Stonehenge and its surrounding landscape were places where the living came to venerate the dead. “I think the people who built Stonehenge brought their ancestors from Wales in stone form,” he says. Last month, his team reported new radiocarbon dates from human cremation burials indicating that Stonehenge had been a “cemetery” back to about 3000 B.C.E., the time of its first construction and before the stones were erected.

Yet the discoveries at Durrington Walls are only relevant to Stonehenge if the two monuments are related. That has been unclear because of uncertainties in dating at both sites. Parker Pearson's team now appears to have closed this dating gap considerably: A reanalysis of earlier Stonehenge radiocarbon dates, published last year in Antiquity, puts the erection of the large stones—which weigh as much as 45 tons and were brought from about 30 kilometers away—at 2600 to 2400 B.C.E. An antler pick apparently used in the construction of the earthwork at Durrington Walls clocked in at 2570 to 2350 B.C.E., and a pig bone there was dated to between 2830 and 2470 B.C.E. “Mike needed to get those two monuments closer together in time, and he's done it,” says Richard Bradley, an archaeologist at the University of Reading in the U.K., though he notes that some wiggle room remains.

Even if Durrington Walls and Stonehenge were part of a single “ritual landscape,” as Parker Pearson has argued, their symbolic meaning is still open to debate. For example, archaeologist Timothy Darvill of Bournemouth University in Dorset, U.K., has coproposed the idea that the bluestones were transported from Wales because of their healing properties and that Stonehenge was a center of healing (Science, 11 April, p. 159). “To have animals coming in from Wales fits our idea quite nicely,” says Darvill.

This summer, Parker Pearson's team plans to find out more about the founders of Stonehenge by excavating possible seasonal houses just west of the monument. The new site is roughly dated to about 3000 B.C.E., when Stonehenge was still just an earthwork circle and the monument we see today was just a gleam in a prehistoric eye.

2. FRANCE

# Despite Protest, CNRS Moves Toward Major Shakeup

1. Martin Enserink

PARIS—Researchers and science labor unions last week stopped a proposed reform of one of Europe's biggest research agencies with their bodies. But their victory may be short-lived, as France's science ministry says the makeover of the National Center for Scientific Research (CNRS) will proceed.

On 19 June, the leaders of more than 1000 protesting researchers and workers occupied an ornate meeting room inside the Parisian headquarters of CNRS, where the agency's board of trustees was to vote on the controversial reform plan. Worried that angry protesters might take over the entire building, CNRS President Catherine Bréchignac canceled the meeting.

The proposed shakeup would create eight new institutes within CNRS. The protesters say that amounts to “dismantling” the science flagship. But other researchers contend that the reforms will improve CNRS and could plant the seeds for a new national institute that would unify the life sciences in France.

With a €3.3 billion budget and a staff of 32,000, CNRS spans the disciplines from anthropology to astrophysics. Its researchers—civil servants with jobs for life—are spread out across the country, often working in close collaboration with university scientists. “Horizon 2020,” a strategic plan drawn up by the center's leadership and France's ministry of higher education and research, proposes to replace CNRS's department-based structure with institutes based on scientific fields such as chemistry, mathematics, and physics. Some of these could become “national institutes” and take a leading role. The structure would be clearer and more efficient, according to the plan.

But the unions and Sauvons la Recherche (SLR), a movement founded in 2003 to protest budget cuts by the previous government, says the plan would be the “death” of CNRS and another step toward a U.S.-style system in which researchers face cutthroat competition and permanent job insecurity. The groups have fought previous reforms such as the creation of a national research agency for project-based funding and a law that gives universities more autonomy (see Editorial, p. 1695).

CNRS's own scientific council isn't happy, either. On 16 June, its members voted 10 to 7 against the plan. The breakup into distinct institutes would hamper interdisciplinary work, a key CNRS strength, says anthropobiologist Gilles Boëtsch, who chairs the council.

Under CNRS rules, the board must hold another meeting within 20 days with the same agenda. Union and SLR leaders want Bréchignac and French science minister Valérie Pécresse to hold a new round of consultations and delay implementation of any reform. A ministry spokesperson says that Pécresse will reassure scientists that CNRS will continue to exist and preserve the rights of staff but that she will stick with the plan.

Complicating the debate are several long-running divisions. Some SLR members accuse President Nicolas Sarkozy's conservative government of settling political scores with the left-leaning CNRS. There have also been tensions within CNRS between physicists, who have dominated its leadership, and biologists, who say they have been given short shrift. (Bréchignac, an atomic physicist, fanned those flames when she said recently that CNRS biologists could perform better.)

Biologist Jules Hoffmann, president of the French Academy of Sciences, says the time is ripe for coordinating programs at CNRS, the National Institute for Health and Medical Research (INSERM), and perhaps other government players. The creation of a strong life sciences institute within CNRS, he adds, could even be the first step toward a full-fledged merger, which could correct a fragmented structure that weakens French science.

3. U.S. OCEAN POLICY

# Proposed Rule Would Limit Fish Catch but Faces Data Gaps

The U.S. government has proposed first-ever annual catch limits in an attempt to stop overfishing.

Environmentalists are welcoming the draft rule, published in the Federal Register on 9 June by the National Oceanic and Atmospheric Administration (NOAA). But experts caution that it will be difficult—and hugely expensive—for the agency to regulate the many marine species about which little is known. Some scientists also wor ry about economic repercussions if the rule ends up curtailing fishing in healthy populations. “It could have staggering consequences,” says fisheries biologist Ray Hilborn of the University of Washington, Seattle.

In December 2006, Congress made extensive changes to the federal law that governs fishery management policy (Science, 22 December 2006, p. 1857). The job of implementing those changes falls to NOAA's National Marine Fisheries Service (NMFS), which manages more than 1000 marine species, not all of which are economically important. Some 41 of the 528 stocks that NMFS monitors are being overfished, mostly off the East Coast.

The rule spells out how NMFS intends to end the overfishing, rebuild depleted stocks, and ensure “optimum yield.” All eight regional fishery management councils would be required to set annual catch limits, which must be approved by a council's scientific advisory committee. The limits must incorporate a safety margin to account for scientific uncertainty surrounding the stock assessment, as well as uncertainty about technical aspects of implementation.

There's tough enforcement language in the new rules. If the councils don't meet their deadline for rebuilding overfished stocks, they will have to cut the annual catch limits. Lee Crockett of the Pew Environment Group in Washington, D.C., calls the language “a pleasant surprise.”

One unanswered question is how to deal with so-called data-poor species. “It's a big black box,” says Andrew Cooper of Simon Fraser University in Burnaby, Canada. He predicts that catch limits will be set low and that the fishing industry will agree to contribute more data and analysis to the agency.

Hilborn worries about the negative impact on trawl fisheries, which scoop up large numbers of a few abundant commercial species but also many low-value species about which little is known. Trawlers could be prohibited from catching anything at all in order to protect data-poor species that may not be in danger. “You're going to give up a lot of fish,” he says. NMFS scientists are working on technical guidance about how to deal with data-poor species.

It won't be cheap to fill in the data with rigorous stocks assessments and marine surveys. Accordingly, President George W. Bush has requested an increase of $8.9 million to NOAA's$31.6 million budget next year for fishery assessments, and Congress seems amenable to the hike. But Hilborn doubts that amount would be nearly enough. “To do it right would take a staggering increase in resources,” he says.

Another regulation will give NMFS more data on recreational fishing, which can rival the impact of commercial fishing in some parts of the country, by creating a registry of saltwater anglers. The agency will accept public comments through 11 August, and the catch-limit rule remains open for public comment until 8 September. The agency hopes to finalize both rules by the end of the year.

4. 2008 U.S. BUDGET

# House Gives $400 Million to Four Science Agencies 1. Jeffrey Mervis Science agencies are barely a footnote in the$186 billion supplemental spending bill to continue funding the U.S. war effort in Iraq and Afghanistan approved by the House of Representatives last week. But the footnote includes a welcome bump-up of $400 million for four agencies whose research budgets were flattened late last year by legislators. “It's not that much money. But as a statement of priorities, we're very gratified,” says Howard Garrison of the Federation of American Societies for Experimental Biology in Bethesda, Maryland, referring to the$150 million that the House approved for the National Institutes of Health for the 2008 fiscal year that runs until 30 September. That could fund 260 additional grants across most of the 27 institutes and centers. Lawmakers also doled out $62.5 million each for the National Science Foundation (NSF), the Department of Energy's Office of Science, and NASA (ScienceNOW, 20 June), plus$62.5 million for DOE's environmental cleanup efforts at Hanford, Washington.

Science advocates have been lobbying for much more—$900 million—for NSF, DOE science, and the National Institute of Standards and Technology to restore those agencies to levels requested by President George W. Bush in his 2008 budget and initially backed by Congress before a last-minute reversal (Science, 4 January, p. 18). Although the White House loudly opposed adding domestic spending to the war supplemental, last week it bent to pressure from House Democrats and agreed to accept expanded unemployment and veterans education benefits as well as$8.5 billion in emergency spending for disaster relief. The Senate, which last month had approved $1.2 billion more for research in its version of the war supplemental, was expected this week to accede to the terms of the House bill (H.R. 2642). The DOE science funding is intended to stave off layoffs at two high-energy physics laboratories, Fermi National Accelerator Laboratory and the Stanford Linear Accelerator Center (Science, 11 January, p. 142). Most of the additional NSF funding will go to improving precollege math and science instruction through its existing Noyce Scholarship program for undergraduates and a new master's level program modeled on a 4-year-old initiative in New York City called Math for America. The NASA funding will bolster science and aeronautics programs cut to fix the shuttle system in the wake of the 2003 Columbia disaster. The$400 million in the supplemental represents what House and Senate Democratic leaders decided they could afford after agreeing to extend a helping hand to the scientific community. “Any split was as rational as any other,” explains a congressional aide about the allocation between NSF and NASA. “I'm not sure the [2008] requests had anything to do with it.”

5. FUSION REACTOR

# ITER Costs Give Partners Pause

1. Daniel Clery

Last week, ITER scientists revealed a new cost estimate for the multibillion-dollar fusion reactor that was 30% higher than earlier calculations. Now the project's seven international partners must decide whether they can afford it.

ITER, or the International Thermonuclear Experimental Reactor, is designed to show conclusively that fusing together hydrogen isotopes at extreme temperatures—the process that powers the sun—can be harnessed on Earth as a practical energy source. Fifteen years of discussion and experiment led in 2001 to a “final” design for the 20,000-ton ITER reactor, twice the size in linear dimensions of the world's current largest. Since then, the partners—China, the European Union, India, Japan, Russia, South Korea, and the United States—have chosen Cadarache in southern France as a site and set up the organization that will build the reactor (Science, 13 October 2006, p. 238).

The current price tag is €10 billion, half of which will pay for construction. Last week, the project's governing council met in Aomori, Japan, to hear about a new review of that 2001 design that includes numerous refinements and upgrades to components, including magnets and heating systems, plus additional magnets to help control explosive discharges at the plasma edge (Science, 13 June, p. 1405). Those design changes will cost an extra €1.2 billion to €1.6 billion, ITER managers estimate, and the council immediately ordered an independent assessment of the costs in time for its next meeting in November. In the meantime, the council did approve a 2-year delay, to 2018, in the expected start-up of the reactor.

Fusion experts say that it's notoriously hard to keep such large projects within budget. “When they actually go out and build things, they always cost more,” says Stephen Dean, president of Fusion Power Associates, a lobby group in Gaithersburg, Maryland. But ITER scientists believe that the design changes are crucial to the project's chance of success and that the partners should approve the new cost estimate. “It will define what we can do and when we can do it,” says David Campbell, assistant head of ITER's department of fusion science and technology. It won't be an easy sell, however: Some ITER partner governments won't be happy at being asked to fork out more.

The panel tasked with assessing the new cost estimate will be led by Frank Briscoe, former operations director of the JET fusion reactor near Oxford, U.K. The European Union, which as host must bear nearly 50% of the cost, declined comment on the new estimate beyond saying, in the words of research spokesperson Catherine Ray, that “we're happy [Briscoe's] group has been set up.” Meanwhile, the partners in the world's most expensive experiment will be debating its future. “There will be some very, very hard diplomatic negotiations over what the partners are prepared to pay,” says a senior European researcher who asked not to be named.

6. ETHICS

# Senate Inquiry on Research Conflicts Shifts to Grantees

1. Jocelyn Kaiser

Senate investigators began poking around academic medical centers last summer, looking for information on who was receiving corporate money and who was reporting it in compliance with conflict-of-interest rules. At the same time, they asked drug companies to name whom they were paying, and how much. This month, the two halves of the bomb came together, revealing discrepancies with a bang. The fallout struck Harvard Medical School and the affiliated Massachusetts General Hospital (MGH) in Boston. Congressional sleuths allege that three faculty psychiatrists failed to properly report hundreds of thousands of dollars of outside income.

As investigators under Senator Charles Grassley (R-IA) sift through the cases, the biomedical community is facing a couple of angst-inducing questions: Will there be more bombshells? And how will grant overseers at the National Institutes of Health (NIH) and academic deans respond? The short answer is, yes, Grassley plans more detonations. The senator has said he is investigating about 30 individuals at 20 universities who may have broken federal conflict-of-interest reporting rules. This week, he asked Stanford University why it did not require that a faculty psychiatrist report the full value of his $6 million in stock in a company that makes a drug being studied in an NIH-funded trial that the psychiatrist oversees. Stanford was preparing a statement as Science went to press (ScienceNOW, 24 June). Universities, meanwhile, say they're scrambling to tighten procedures to track conflicts, hoping to reassure the public and stave off more stringent measures that they say could stifle cooperation with industry. “I think the community's been awakened,” says David Korn, former dean of Stanford University School of Medicine and now a senior vice president of the Association of American Medical Colleges (AAMC) in Washington, D.C. Health policy researcher Eric Campbell of MGH, who has documented the prevalence of industry ties in academia, says an overhaul is long overdue: “Consulting has been one of the great wink-winks of all time.” Biomedical researchers in government came under pressure 4 years ago, when the House Energy and Commerce Committee investigated media reports that several top-level NIH intramural researchers had failed to report consulting income. That led NIH Director Elias Zerhouni to ban all company consulting by in-house researchers. Now NIH-funded extramural researchers are feeling the heat, thanks to Grassley, ranking minority member of the Senate Finance Committee. A 1995 Public Health Service (PHS) regulation requires that investigators seeking NIH funding tell their institutions about income of$10,000 or more a year, or 5% equity in a company, “that would reasonably appear to be affected by the research.” Last year, Grassley found that child psychiatrist Melissa DelBello of the University of Cincinnati in Ohio had apparently underreported $138,000 in drug company income. Earlier this month, Grassley netted the Boston fish: Joseph Biederman, Timothy Wilens, and Thomas Spencer, all highly respected child psychiatrists at Harvard and MGH. Grassley says reports filed by the three initially suggested that they had earned about$200,000 each in consulting income since 2000. But they later said the total was actually $3.6 million. When Grassley compared data from drug companies with their conflict-of-interest reports, he found the amounts didn't always match up, according to tables he released. Harvard and MGH have said they are investigating. Grassley's tables suggest a precision that may not exist in academic forms. For example, Harvard and MGH required only ranges of income to be reported, not specific amounts. In some cases, the Harvard or MGH investigators overreported income from a company one year, suggesting a mismatch in dates. Other discrepancies make them look more culpable, however. Two of the investigators led NIH-funded clinical trials for psychiatric drugs but may have failed to report income from the drugs' makers that exceeded the$10,000 PHS threshold—if so, a clear conflict.

If any of the three have indeed broken the PHS reporting rules, Harvard and MGH could be subject to fines or suspension of grants, say NIH officials. NIH says it is undertaking a systemwide review of its conflict-of-interest policies, after the Department of Health and Human Services inspector general hammered the agency for lax oversight in January (ScienceNOW, 18 January). However, NIH has rejected a recommendation that it routinely collect details from institutions on their management of conflicts. That responsibility should remain with institutions, says NIH extramural research chief Norka Ruiz Bravo.

Institutions acknowledge that they have a long way to go. AAMC and the Association of American Universities (AAU) have issued conflict-of-interest guidelines, most recently in February. They recommended that all payments, not just those above the $10,000 limit, be reported, and in specific amounts. But only some institutions follow these policies now. And many are still migrating from paper forms to electronic databases that can be easily accessed, say, by ethics review boards. “It has to be electronic to be effective,” says Robert Rich, dean of the School of Medicine at the University of Alabama, Birmingham, and co-chair of the recent AAMC/AAU report. Even then, the best system won't keep some researchers from failing to report. Another idea is for universities to crosscheck faculty reports with a public database. Grassley has proposed a law that would require drug companies to report all payments of more than$500 made to doctors.

The academic community is concerned: “The risk is that we'll have a set of rules imposed on us just at a time when universities should be making sure discoveries are turned into useful products. That would be a catastrophe for America,” says Harvard University Provost Steven Hyman. “By the same token,” he adds, “it is very important that the physician's only interest aligns with the [patient's] interest and not with a company's.” But convincing Congress that universities can “thread the needle” on their own, as Hyman says, will be a challenge.

7. VIROLOGY

# 'Biased' Viruses Suggest New Vaccine Strategy for Polio and Other Diseases

1. Martin Enserink

They're called silent mutations, but they could make a big noise in the vaccine field. Introducing hundreds of these seemingly inconsequential changes into a poliovirus can cripple the virus enough to make it work as a live vaccine in mice, scientists report on page 1784. The technology might lead to safer polio vaccines and perhaps to so-called live attenuated vaccines against other diseases. “It's a nice study and a very promising technology,” says Rino Rappuoli, global head of vaccine research at Novartis.

The new vaccine strategy exploits the fact that almost all amino acids can be encoded by multiple codons, triplets of the DNA bases guanine (G), adenine (A), cytosine (C), and thymine (T). GAA and GAG, for instance, both represent the amino acid glutamic acid. But many organisms, including viruses, have a bias toward certain codons in their genes. This may be because those codons are easier to translate at ribosomes, the cell's protein factories, thus speeding up protein production. Researchers have learned to take advantage of codon bias: For example, they insert a microbe's favorite codons when engineering it to make a desired protein.

The vaccine approach does the opposite: It creates underperforming viruses by giving them unfavored codons. The hope is that these strains will be too weak to cause disease yet produce the same proteins and elicit the same immune response as wild-type virus. Two teams have been exploring that idea independently, by synthesizing a new version of the poliovirus's capsid gene with hundreds of small changes and stitching it into the RNA genome of either a wild poliovirus or a vaccine strain. In 2006, a group led by Olen Kew of the U.S. Centers for Disease Control and Prevention (CDC) in Atlanta, Georgia, first showed that such a strain could barely grow in human cells. Shortly after, a team led by Eckard Wimmer at Stony Brook University in New York state presented a similar study.

Wimmer's group now shows that codon-crippled viruses can work as a vaccine. Rather than directly using codon bias, the team took advantage of a related phenomenon called codon pair bias: that is, codons tend to be followed much more often by certain codons than others. (For instance, the amino acid pair alanine-glutamic acid is much more often encoded by the codons GCA-GAG than by GCC-GAA.) With the help of computer scientists, the team assembled two polioviruses, each of whose capsid protein genes had hundreds of underrepresented codon pairs. Mice injected with either strain didn't get sick but developed immunity against a challenge with a lethal dose of the real poliovirus.

Although a massive worldwide campaign with a live virus vaccine has all but eradicated polio, there's room for a new vaccine. After eradication is complete, plans call for the use of killed vaccines. Production of killed vaccines is risky, because viruses can escape from vaccine factories before they're inactivated. The weakened viruses made by the Stony Brook team could provide a safer starting material for a killed vaccine, says Roland Sutter of the World Health Organization in Geneva, Switzerland—although other potential routes are under study as well. Steffen Mueller, the paper's lead author, says his team hasn't yet killed its polio strain to see if it also works as a vaccine, but there's no reason to believe it wouldn't.

The “big payback” from the new method may be that it can weaken almost any other virus as well, Sutter says. “The nice thing is that we don't really have to understand the virus and the way it functions,” Mueller says. Getting approval for vaccines created this way will be a “huge job,” Rappuoli predicts, because researchers will have to show that the wimpy strains can't mutate and cause disease—the big worry with any live-attenuated vaccine. But the fact that so many mutations are involved should minimize chances of this happening, says Mueller.

8. INFECTIOUS DISEASE

# An Ill Wind, Bringing Meningitis

1. Leslie Roberts

Crippling epidemics of meningococcal meningitis sweep across Africa with the onset of the dry season and harsh harmattan winds. An affordable, effective vaccine in the works could change that.

Crippling epidemics of meningococcal meningitis sweep across Africa with the onset of the dry season and harsh harmattan winds. An affordable, effective vaccine in the works could change that

OUAGADOUGOU, BURKINA FASO; BAMAKO, MALI; GENEVA, SWITZERLAND—The dust is inescapable, burning your eyes, clogging your nose, penetrating into your lungs, and making breathing ragged. In March, on the road to Koudougou, some 100 km west of Ouagadougou, the landscape is moonlike. In the cratered bottom of a lakebed, dust-caked men, barely distinguishable from their surroundings, fashion bricks from the mud. The bricks will dry quickly in the baking heat, which tops 45°C each day.

It is the dry season in Burkina Faso. And with the dust and the hot, dry wind, known as the harmattan, that blasts across the Sahel come meningococcal meningitis epidemics, caused by the bacterium Neisseria meningitidis. What, exactly, about these conditions triggers the epidemics remains mysterious, but they come like clockwork, hitting Burkina Faso every year and engulfing the entire “meningitis belt,” which runs from Ethiopia in the east to Senegal and The Gambia in the west, every 6 to 12 years.

The last big one, in 1996–97, sickened hundreds of thousands and killed more than 25,000 in 10 countries. In 2007, the death toll climbed alarmingly high again, prompting the World Health Organization (WHO) to warn that another huge epidemic was likely in 2008. But this season turned out to be relatively quiet, with some 9400 cases in Burkina Faso and 27,000 across the entire belt. As always, the epidemics in Burkina Faso stopped suddenly with the first rains in May, as the population in this country, one of the poorest in the world, braced for the inevitable onslaught next year.

Koudougou district officially passed the epidemic threshold in mid-March, and scarce supplies of the meningitis vaccine were made available to try to curb the epidemic's spread. At a rudimentary health center there, hundreds of people—mostly women and children—queue up for vaccinations, seeking shade by the buildings or under a scrawny tree. Most have been waiting patiently for hours, but some occasionally surge to the front of the line only to be pushed back by the men in charge of crowd control.

At best, this reactive vaccination strategy, as it is called, is a “Band-Aid,” says Rosamund Lewis, a physician and meningitis expert at the GAVI Alliance (formerly the Global Alliance for Vaccines and Immunization). The reason is that the vaccine being used, a 1960s design using a polysaccharide from the bacterium's coat and still the only affordable one in Africa, doesn't work very well. Although this vaccine prevents those carrying the bacterium from getting sick, it doesn't stop them from passing it on to others; immunity lasts only a few years; and the vaccine has minimal effect on children under age 2. Because of these limitations, WHO has long recommended that it be used only to control epidemics, not to prevent them—a strategy that has its critics. “The epidemic is sometimes over by the time vaccine arrives,” concedes William Perea, a Colombian-born epidemiologist who leads Epidemic Readiness and Interventions at WHO and who nonetheless supports the strategy for lack of a cost-effective alternative.

F. Marc LaForce wants to change all that. He is heading an innovative public-private partnership known as the Meningitis Vaccine Project (MVP) to develop an affordable, effective, long-lasting vaccine for African meningitis—a conjugate vaccine that includes a protein to boost the immune reaction. The conjugate will cost roughly 50 cents a dose—a price many African governments say they can afford—and is already being tested in clinical trials at several African sites. Barring any further delays, it will be introduced in a massive trial of some 9 million people in Burkina Faso in late 2009. LaForce hopes the new vaccine will eventually be used in preventive campaigns across the entire meningitis belt and spell the end of these devastating epidemics.

“It will change completely the approach to meningitis, and that will be great,” says Myriam Henkens, a physician at Médecins Sans Frontières (MSF) in Brussels, Belgium. “We are counting on the conjugate,” she says. “It will be so much cheaper and so much better.”

And all indications are that countries will be clamoring for it. Although other diseases exact a bigger toll, they are not as feared as meningitis, which can kill within 24 hours and often leaves survivors deaf or otherwise disabled. “Boy, do they want this vaccine,” agrees Emil Gotschlich of Rockefeller University in New York City, who developed the existing polysaccharide vaccine.

But the road between here and there is littered with potholes. MVP, a joint project of WHO and the Seattle, Washington-based health nonprofit PATH, has already encountered unexpected obstacles. Originally promised for 2007, the vaccine may not be ready for the 2009 rollout in Burkina Faso, LaForce admits. And questions remain about just how long immunity will last, whether a booster will be needed—which would affect the overall cost of a vaccination strategy—and perhaps more worrisome, whether once N. meningitidis group A—which causes the majority of epidemic disease in Africa and is the target of this vaccine—has been beaten down, other strains will arise to replace it.

“Vaccine-making is not for the faint of heart,” concedes LaForce.

## Mysterious cycles

Meningitis is an infection of the meninges, the thin membrane that surrounds the brain and spinal column. Several different bacteria can cause meningitis, including Haemophilus influenzae type B (Hib) and Streptococcus pneumoniae—and also a few viruses—but only N. meningitidis, or meningococcus, spawns the huge epidemics that sweep across the belt. Of the dozen or so meningococcal groups, A is by far the worst, causing roughly 80% to 90% of epidemic disease most years. (Groups C and B, by contrast, fell adolescents and college students in Europe and the United States.)

Roughly 1 in 20 people “carry” the bacterium asymptomatically in the back of their throats and can transmit it to others; what triggers invasive disease remains mysterious. Once the disease begins, with its characteristic sudden fever, headache, and stiff neck, it progresses rapidly. Even with prompt treatment with antibiotics, which is often impossible in remote villages, about 10% die, and up to 25% of survivors are left with permanent disabilities such as deafness or mental retardation.

For about 100 years now, experts have puzzled over the remarkable seasonality of the epidemics, which begin with the dry season in December, peak after a few months, and disappear with the first rains in May or June. One hypothesis is that dust and wind increase transmission of the disease, which is spread person to person through respiratory droplets. The other, which seems more likely, is that the harsh environmental conditions irritate the mucus membranes, enabling the bacterium to more easily penetrate and enter the spinal fluid, where it causes invasive disease.

Similarly, although epidemics are inevitable, almost all efforts to predict exactly when and where they will strike have failed miserably. It has defied the best scientific minds, says Perea, who also works with MVP.

## Failed policy?

MVP was born of what everyone refers to as the “terrible epidemic” in 1996–97. African leaders and global health officials watched in horror as the largest meningococcal meningitis epidemic ever recorded swept across the belt. More than 250,000 people fell ill, and the death toll soared beyond 25,000. Clearly, the epidemic response system wasn't working.

Shortly thereafter, WHO, UNICEF, MSF, and the International Federation of Red Cross and Red Crescent Societies banded together to create the International Coordinating Group (ICG) on Vaccine Provision for Epidemic Meningitis Control. Headquartered at WHO in Geneva, Switzerland, ICG tries to ensure that limited supplies of the polysaccharide vaccine are rapidly sent to where they can do the most good. Following the recommendations of a pivotal 2000 Lancet paper by Lewis—then at MSF's research institute, Epicentre—and colleagues, WHO lowered the epidemic threshold to 10 cases among 100,000 people in 1 week, shaving more than a week off the usual response time. Perea says that if a country can launch a reactive vaccination campaign within 3 to 4 weeks of an epidemic's onset, it can prevent 70% of the cases. Response times have improved substantially, but campaigns start within 3 weeks only about 60% of the time, he says.

These improvements were just a stopgap, all conceded. What was needed, recommended a WHO advisory group in 2000, was an entirely new type of meningitis vaccine, a conjugate vaccine that would confer lasting immunity and could be used preventively, modeled on the Men C vaccine that had all but eliminated the disease in the United Kingdom.

A conjugate vaccine uses the same polysaccharide but links, or conjugates, it to a protein to increase its immunogenicity. Conjugate vaccines developed to date, including those for Hib and S. pneumoniae, confer longer lasting immunity than polysaccharide vaccines, work in infants, and, perhaps more important, reduce transmission of the bacteria, thereby providing herd immunity and protecting even those who are not immunized.

The newly established Bill and Melinda Gates Foundation didn't need much convincing; in May 2001, it sprang for $70 million over 10 years to establish MVP to develop, test, license, and introduce an affordable conjugate vaccine for Africa. From the outset, some questioned the approach, worrying that it would take too long and arguing that alternatives already existed. Gotschlich and vaccinologist John B. Robbins of the U.S. National Institute of Child Health and Human Development, who both won Lasker Awards for their pioneering work, had long railed against what they considered WHO's failed policy of waiting for an epidemic before starting mass vaccination. In a series of papers, including a roundtable in the WHO Bulletin in 2003, they argued that WHO should urge countries to use the polysaccharide vaccine preventively. True, the polysaccharide vaccine is by no means perfect, says Gotschlich—although he says it is more effective than others now acknowledge. And without question, he says, the conjugate vaccine will be far superior: “I am all for it,” says Gotschlich, who serves on MVP's advisory committee. But the polysaccharide vaccine was available and could save lives right away for just pennies a dose, says Gotschlich, who kicks himself for not urging WHO to be proactive back in the 1970s. “In theory, he is right,” says Perea, but he doubts the strategy would work. And it would be hugely expensive. Because epidemics are so unpredictable, he says, this approach would require vaccinating the entire population of the meningitis belt—roughly 400 million people—every 3 years for an estimated$400 million a campaign. “To arrange a countrywide vaccination every 3 years is nuts. The number of resources, from my point of view, is totally unjustified to introduce the polysaccharide as a preventive vaccine,” says Perea. With limited resources, he asks, why not focus instead on a “real solution”: a conjugate vaccine.

## Trials and tribulations

Two months after the Gates money came through, LaForce was on the job, setting up shop in July with a small, energetic staff in Ferney-Voltaire, across the French border from WHO headquarters in Geneva. An infectious disease expert and former meningitis officer for the U.S. Centers for Disease Control and Prevention (CDC), LaForce had recently quit academic medicine, fed up with the paperwork and the ever-increasing administrative demands. “I wanted to make an impact on global public health before I retired,” he says.

In August, LaForce and WHO's Luis Jodar, an early member of MVP, were on a plane to Africa where they began informal consultations with African leaders and health officials about what they wanted in a vaccine. He distinctly remembers a discussion with Hassane Adamou, the secretary general for Niger's Ministry of Health, who said: “Please don't give us a vaccine we can't afford. That is worse than no vaccine.” When he asked leaders what was affordable, the answer was 50 cents a dose or less, significantly lower than the $2 to$3 a dose the collaborators had originally envisioned. Won over, LaForce immediately began pushing for the best and least expensive vaccine possible: a First World vaccine at a Third World price.

For at least 100 years, when the hot, dry harmattan wind blows, meningitis outbreaks have swept across the meningitis belt (see map). Experts suspect that the unforgiving environmental conditions make the body more susceptible to the bacterium. When meningitis cases soared in 2007, WHO warned that 2008 might be an especially bad year, but it was relatively mild. No one knows what 2009 will bring.

Vaccine expert Rino Rappuoli, now head of global vaccine research at Novartis, argued vociferously that waiting for the cheapest vaccine was actually more expensive, in terms of lives lost in the interim. While at Chiron in the 1990s, Rappuoli had developed the first conjugate vaccine against meningitis, targeted against groups A and C. It was a “beautiful vaccine,” he laments, that performed well in field tests in Niger and The Gambia. But as the group C component was developed for the lucrative market in the United Kingdom, “we were asked to remove A,” which was needed only for Africa, he recalls. “The market was only for C.” Instead of reinventing the wheel, urged Rappuoli, MVP should dust off his vaccine and make it quickly available.

MVP stuck to its plan for an affordable vaccine. But LaForce says he could find no major vaccine manufacturer willing to produce a Men A conjugate for $2 a dose, much less the price he thought was needed. So in an unusual strategy, LaForce insisted on a guaranteed selling price—50 cents or less—and then found a developing country manufacturer, Serum Institute of India Limited (SIIL) in Pune, willing to take it on. (Although the company has agreed to a fixed price for Africa, it is free to sell the vaccine elsewhere at higher prices.) Next, MVP lined up suppliers for the raw ingredients: the group A polysaccharide and the protein it would be conjugated with, a tetanus toxoid. And they contracted with a European research group to develop a new conjugation technology and then transfer it to SIIL. From the outset, MVP decided to concentrate on a monovalent vaccine against group A. Creating a multivalent vaccine that could protect against the other groups in Africa would jack up the cost, not to mention the risk of failure, and stretch out the time frame of development. “If we could [address] 85% of the burden with the simplest approach, … to me, that was a completely acceptable wager with public money,” says LaForce. The unexpected emergence of a new epidemic strain, W135, in 2000 caused considerable soul-searching. Even though W135 appeared to be a significant new threat, in the interest of speed, MVP decided against a midcourse change. “History will tell if that was the right decision,” says Lewis. Manufacturing didn't go smoothly. The conjugation technology proved finicky, causing some delays. But still, the partners thought they were roughly on track until, in spring 2003, the European research group MVP was collaborating with announced it was unwilling to transfer the technology to SIIL. That was the absolute low point, says LaForce. Even friends of the project said it was doomed, he recalls, and several called for his ouster. But others lobbied to give him more time. MVP found a solution at the U.S. Food and Drug Administration, where longtime vaccine experts Robert Lee and Carl Frasch at the Center for Biologics Evaluation and Research had already developed an alternative conjugation technology and quickly transferred it to SIIL with no strings attached. “These guys are heroes,” says LaForce. Preclinical animal studies began soon after. Since then, LaForce has been traveling around the world selling his vision—a continent free of deadly meningitis outbreaks—while generally greasing the wheels for the vaccine's introduction. LaForce, a 69-year-old American, is a big man with a big voice. “Mon ami, mon ami, comment ça va?” he boomed on a recent trip to Ouagadougou, as he clasped hands and patted shoulders of collaborators and hotel clerks alike. “Le vieux blanc,” or old white man, as he sometimes refers to himself, is invariably upbeat, even as he delivers the bad news that, because of a regulatory snafu in India, the vaccine's introduction in Burkina Faso will be delayed from 2008 until 2009. But the vaccine will come, he assures, and it will be great. ## Thinking locally Phase II and II/III clinical trials of MVP's Men A vaccine candidate are under way in Mali, The Gambia, Senegal, and India, where SIIL plans to license the vaccine to protect against that country's occasional outbreaks. Samba Sow, a Malian physician and epidemiologist, is running the trials in Bamako, Mali, at the Center for Vaccine Development, a partner lab of the University of Maryland's CVD. With MVP support, a former leprosarium—a small group of former patients still lives on the grounds—has been converted into a cheery clinical center for testing the new conjugate and other vaccine candidates. The challenge in setting up the trials in some of the poorest countries in the world, says Simonetta Viviani, an Italian physician who heads MVP's vaccine development from Ferney-Voltaire, was to create a “functional but minimalist” system that would meet all international standards for ethics and good clinical practices but could also be continued with local experts once MVP is gone. That means hiring and training local staff, working closely with the community, and respecting local traditions, she says (see sidebar). In Mali, explains Sow, the initial contacts with the community—and also the first person parents see when bringing their children to the clinic—should be older and preferably religious, which implies a certain wisdom and trustworthiness. Even in this predominantly Muslim nation, says Sow, a Catholic nun still carries significant clout, whereas bright, young doctors, no matter how prestigious their degrees, or white people will not pass muster, he says. Results from the “pivotal” phase II trial of the Men A conjugate vaccine, conducted here and in Basse, The Gambia, and announced in June 2007, “put us on the map in Africa,” says LaForce. The trial of 600 healthy toddlers age 12 to 23 months, half at each site, showed that the conjugate vaccine produced antibody titers almost 20-fold higher than the current polysaccharide vaccine. MVP's clinical team recently unblinded the results from the second arm of the study, in which the same cohort of 12- to 23-month-olds were randomized to receive a booster dose of the conjugate or the polysaccharide or the control vaccines 8 to 12 months later. The as-yet-unpublished data are “fantastic,” raves Viviani. She and LaForce suspect that the vaccine will protect for at least 10 years, although Gotschlich and others caution that the duration of protection won't be known until the vaccine is used in real-world conditions. A phase II/III trial of 900 participants and controls, under way in Bamako, Basse, and Dakar, Senegal, is testing the safety and immunogenicity of a single dose in 2- to 29- year-olds and will also look at its effect on “carriage”—that is, whether it actually does reduce the load of bacteria carried in the back of the throat. Next up is a study of safety and immunogenicity of different dose schedules in infants, expected to start later this year in Ghana. ## Optimism, tempered If the remaining trials go as expected, if the lot-consistency studies under way in Pune, India, go without a hitch, if production can be scaled up, if India licenses the vaccine and WHO “prequalifies” it, if funding comes through, and tens of other details go right, MVP will introduce the vaccine in Burkina Faso in 2009 or perhaps 2010. And that, LaForce hopes, will be the beginning of the end of meningitis epidemics in Africa. The vaccine will be given to Burkina Faso's entire population of 1- to 29-year-olds, roughly 9 million people. Kader Konde, director of WHO's Multi-Disease Surveillance Centre (MDSC) in Ouagadougou, who is also MVP's general troubleshooter for Africa, says the president and senior health officials are on board and are pushing MVP to move faster. LaForce says he wishes they could but adds that “it's important that all the regulatory steps are taken so no one feels they have a substandard product.” MVP has already lined up partners to conduct follow-on studies to measure the vaccine's impact. Surveillance will be critical. CDC will help MDSC look for changes in circulating strains. “We must be able to document any case to see if it is a failure of the vaccine or another strain,” like W135, rearing its head, says MDSC epidemiologist Mamoudou Harouna Djingarey. That requires strengthening surveillance across the entire belt. The hub of these efforts is MDSC in Ouagadougou. Housed in a building that still bears the name Onchocerciasis, the revamped facility boasts state-of-the-art equipment, including a real-time polymerase chain reaction machine, a recent gift from CDC, for analyzing cerebrospinal fluid samples to determine the bug and the group. Meanwhile, a half-dozen epidemiologists, microbiologists, and data experts track the bug's every move in 14 countries. Before MDSC, there was “not much,” recalls Perea, who notes that surveillance is now “pretty good” but is still spotty in some countries, such as Chad and Nigeria. Even in Burkina Faso, adds Djingarey, cases are still missed, cerebrospinal samples are degraded in transport, and data dribble in late from some districts. All that must be fixed, he says. Following the planned 2009 introduction, MVP, MDSC, the Burkina Faso Ministry of Health, and CDC, in collaboration with the Norwegian Institute of Public Health and the Centre for Prevention of Global Infections at the University of Oslo, will conduct carriage studies. Other academic, governmental, and nongovernmental partners, coordinated by Brian Greenwood at the London School of Hygiene and Tropical Medicine, will monitor how long immunity lasts and whether a booster shot is needed, as turned out to be the case with the Men C conjugate in the United Kingdom. Provided no major problems surface, WHO and its AFRO bureau and UNICEF will introduce the vaccine, first in the three hyperendemic countries: Burkina Faso, Mali, and Niger. Because production will be limited to about 45 million doses for the first few years, the partners are trying to allocate it to the populations at highest risk. By 2016, there should be enough vaccine for the most vulnerable population of the meningitis belt, roughly 250 million people, says LaForce. He envisions that countries will do “catch-up” vaccination campaigns every 5 years or so until the vaccine is approved for infants and can be integrated into routine childhood immunizations. As Science went to press, the GAVI secretariat recommended that its board approve$370 million to cover the vaccine introduction in Burkina Faso and subsidize the vaccine's rollout across the belt. Eventually, GAVI's support would wane and countries would pick up the tab themselves.

In the interim, stresses Perea, it will be essential to keep up supplies of the polysaccharide vaccine, which plummeted because of a production decline after MVP was announced, leaving a global shortfall that still persists. It was an unpredictable market to begin with, and manufacturers thought the conjugate would “put them out of business,” says LaForce, so “most moved on.”

“They stopped 10 years too soon,” says Lewis. Since then, two manufacturers have agreed to produce the vaccine if WHO guarantees to purchase it.

The next priority, all agree, is an affordable multivalent conjugate vaccine that would offer even wider protection from all meningococcal groups in Africa. (The Menactra quadrivalent conjugate vaccine licensed in the United States sells for about $100 a dose.) Right now, it's not clear who will take the lead on the multivalent. MVP won't, says LaForce—although he hopes it has shown what is possible—as the project will shut its doors and he will retire in 2011. The rains started in late May in Burkina Faso, and the epidemic, which had affected more than 9000 and killed 900 there, waned. Although that's a much lower toll than everyone had feared for this year, “it's still 9000 cases too many,” says LaForce. He won't hazard a guess about how bad next year's epidemic will be, but he is hoping for a quiet season. “We need another year's breathing room,” he says, before the next big one hits. 9. INFECTIOUS DISEASE # Costs of Meningitis Outbreaks Are Crippling, Too 1. Leslie Roberts One illness in a family can exact a huge toll on household income, not only in direct costs but indirect ones as well, including loss of income and property such as cattle and crops. OUAGADOUGOU, BURKINA FASO—The government of Burkina Faso prides itself on providing free health care to anyone affected by meningitis. Try telling that to the family from Koudougou gathered at Yalgado Hospital here in the capital city. Last year, during a “reactive” vaccination campaign, their daughter hid from the vaccinators, afraid of the needle, her mother explains. That may be why she got sick this March when another epidemic hit the same district. Now the 14-year-old is lying on a bare cot in a sparse, concrete-floored room, naked in the heat except for her underpants and an IV dripping into her thin arm. She has been here 2 weeks and still has a stiff neck and seems listless. Rigobert Thiombiano, the head of the infectious disease ward, suspects she may have septicemia. Both her parents are with her in the hospital room, as is the custom in Burkina Faso. Hospitals here do not provide the services Western patients take for granted. The family must buy the medicine, provide and cook the food, wash the laundry, and otherwise care for their kin. One illness in a family can exact a huge toll on household income, says Anaïs Colombini, a health economist at the French aid group Agence de Médicine Préventive (AMP) in Ouagadougou, who, with her colleagues, recently completed a detailed socioeconomic study, supported by the Meningitis Vaccine Project (MVP) and the World Health Organization, of the burden of meningitis in Burkina Faso. Government claims aside, AMP found that most families pay on average$25 in direct medical costs, which includes medicine, testing, and lab analysis. Other nonmedical direct costs, say, for food, soap, transportation, and telephone, run another $15. The indirect costs, from loss of income and property such as cattle and crops, are even higher, running$50 per episode. That adds up to almost half of a family's average annual income of roughly \$220. Worse still, says Colombini, just one episode can throw a family into “a downward spiral of poverty” from which it can be impossible to recover.

“Most people don't realize that meningitis is such an important contributor to poverty,” she says, with both short- and long-term costs. “For instance, if you can't be home to tend to the crops, they die, and future income is lost.” Colombini is eager for MVP's new meningitis vaccine, which she hopes will help eliminate poverty along with this feared disease in Africa.

10. INFECTIOUS DISEASE

# Clinical Trials: Dispelling Suspicions, Building Trust in Mali

1. Leslie Roberts

Across West Africa, suspicions of Western medicine--and in particular the fear of being used as a guinea pig in clinical trials--run high. So winning the trust of the local community to enlist participants in clinical trials and ensure that consent is truly informed is a task that can't be undertaken lightly.

BAMAKO, MALI—Across West Africa, suspicions of Western medicine—and in particular the fear of being used as a guinea pig in clinical trials—run high. That is true here in Bamako, where Samba Sow of the Center for Vaccine Development (CVD) is doing clinical testing of a new conjugate vaccine developed by the Meningitis Vaccine Project (MVP) (see main text).

Before trials can begin, Sow, like his counterparts leading trials in The Gambia and Senegal, has to win the trust of the local community so he can enlist participants and ensure that consent is truly informed—a task that can't be undertaken lightly, Sow says. He does it by convincing the poor, largely Muslim and illiterate or semiliterate population that the vaccine being tested is designed to help break the cycle of deadly meningitis epidemics rather than make them sterile or infect them with the AIDS virus, as is widely believed.

At the same time, he and MVP have to design a study protocol and consent process that will pass muster with the four institutional review boards (IRBs) that oversee the studies: one at each of MVP's partners, PATH in Seattle, Washington, and the World Health Organization; one at the University of Bamako; and another at the University of Maryland (UMD), with which CVD-Mali is affiliated. A Malian and Muslim physician who went to medical school in Mali and then was trained in epidemiology at the London School of Hygiene and Tropical Medicine—and who now also serves on the faculty of UMD—Sow manages to bridge both worlds.

Sow starts with the all-important “chef de village”—Bamako, with a population of 1.5 million to 2 million, has six districts, which in turn contain five to 12 “souscartiers,” or local jurisdictions, each with its own leader. Sow explains the trial to the village head and his advisers, showing them the regulatory approvals from the Minister of Health and the Bamako regional health director, and ethical approvals from all four IRBs, and asks him to convene a community meeting.

“If the leader is not convinced, they will say no, and no one is allowed to sign up. That is the way it works in Mali,” Sow says. On the flip side, he says, once one leader says yes, others are also likely to. Sow chairs every single community meeting, during which he describes the study and how the vaccine “acts like a soldier in the body” to protect against meningitis; then the community can ask questions.

The information sheet and consent forms are available in print and on audiotape in English, French, and the most common spoken language here, Bambara. Some questions concern the motives of his group and the Western collaborators. He answers: “Why would I hurt the people where I grew up and who paid for my education? I am here to make my country proud.” Other questions are specific, such as “Why do you have to take blood, and why twice?”

The process is repeated again when individuals sign up. Each participant—or the mother if the participant is a child—meets privately with a physician at CVD's clinical center, where they again review the forms and listen to the audiotape. Sometimes the mothers go home to confer with family and community before deciding, says Sow. But so far, the refusal rate is “very, very low.”

11. EVOLUTION

# Building the Tree of Life, Genome by Genome

1. Elizabeth Pennisi

Cheaper sequencing has put many more genes into the hands of researchers trying to sort out the degree of relatedness of a menagerie of organisms. Thanks to one such "phylogenomic" analysis reported on page 1763 of this week's issue of Science, bird guides may never be the same.

Cheaper sequencing has put many more genes into the hands of researchers trying to sort out the degree of relatedness of a menagerie of organisms

Phylogenetic studies have gone ‘omic. Whereas researchers used to be satisfied comparing one gene, or a few, to sort out the branching of the tree of life, the push now among those building phylogenies is to consider whole genomes—at the very least, dozens of genes and thousands of DNA bases—in establishing kinships among flora and fauna. In this way, evolutionary biology is joining the bandwagon of data-intensive studies pioneered by genomics.

Thanks to one such phylogenomic analysis reported on page 1763, bird guides may never be the same. According to this new avian family tree, grebes will share a section with flamingos, not loons. Dull brown night jars and iridescent hummingbirds would now go together. Even parrots and songbirds share a closer kinship than has been appreciated, says Shannon Hackett, an ornithologist at The Field Museum of Natural History in Chicago, Illinois.

She and more than a dozen colleagues constructed the new genealogy after analyzing 32,000 bases from 19 genes in 169 species. More than just rearranging which birds perch on what branches of the tree, the results raise questions about the evolution of flight; some birds that don't fly are unexpectedly grouped with those that do. “It's the most impressive paper in the higher level phylogeny of birds to come along in a long time,” says Joel Cracraft, an evolutionary biologist at the American Museum of Natural History in New York City. “It will be used by avian systematists and non-avian systematists for a very long time.”

The bird work follows two phylogenomic studies published over the past 3 months that have shaken up perceived evolutionary relationships among animals and more broadly among eukaryotes. In the former effort, a team led by Casey Dunn, now at Brown University, has rearranged the animal kingdom such that comb jellies, not sponges, are among the earliest fauna. In the latter, a European team now divides eukaryotes into two megagroups, not a half-dozen. Together, the three trees speak to the potential of phylogenomics. “We are just beginning to understand what large sequence data sets have to say about the evolution of life on Earth,” says Hackett.

## Entering the genome age

The term “phylogenomics” was coined by Jonathan Eisen a decade ago to describe incipient efforts to integrate evolutionary thinking into genomic analyses and vice versa. What this evolutionary biologist at the University of California, Davis, had in mind was using information about the relatedness of newly sequenced organisms to help sort out gene function and identify comparable stretches of DNA in genomes that have been deciphered. But the term has been “kidnapped,” says Eisen jokingly, by the likes of Hackett and others to describe large-scale efforts to build family trees based on lots of molecular data.

Systematists may like the label, but there's no agreement about how many genes it takes to make an evolutionary tree phylogenomic. “We would say our study is phylogenomic because we have sampled many different genes from many different chromosomes across a subset of avian species, but others would say we still sampled a small portion of the genome,” Hackett points out. And ornithologist Michael Sorenson of Boston University applies an even tougher standard: “I would reserve the term for what lies ahead, i.e., comparisons of whole genomes.”

In traditional molecular phylogeny, researchers pick out a short stretch of one gene, often a mitochondrial gene, count up the sequence differences between species in that stretch, and use sophisticated computer programs to come up with the hierarchy of evolutionary relationships between the species. Most simply, the fewer the differences, the more closely related two species were considered to be.

Gradually, however, researchers realized “that single-gene trees are prone to errors and that many genes are necessary,” explains Jose Castresana of the CSIC Institute of Molecular Biology of Barcelona in Spain. Because genes can evolve at different rates, it's not always possible to pinpoint the true time a species under consideration diverged from a common ancestor by looking just at the changes in one gene from that species. In some cases, there are too few changes to provide statistically reliable results. Other times, the transfer of a gene from one species to another causes phylogenetic chaos.

When Hackett, her postdoc Sushma Reddy, Rebecca Kimball of the University of Florida, Gainesville, and colleagues started their avian project in 2003, collaborators first did a computer simulation to determine how much and what kind of DNA sequence would enable them to figure out the early history of birds. The simulation directed the team to collect at least 20,000 bases from introns and intergenic regions, where mutations occur frequently enough for there to be significant differences in the various lineages. At first, the researchers sampled only about 75 species, but after realizing how much more robust results would be with a larger number, they doubled it.

“The ultimate goal is to provide the rest of the ornithological community with the roots and base of the tree that they can leaf out more effectively,” Hackett says. Traditionally, avian systematists have had trouble sorting out those early days of bird evolution, notes Harvard University ornithologist Scott Edwards. The new results are “bold in setting an agenda for future research,” he says.

In agreement with previous avian phylogenies, Hackett, Kimball, and Reddy found that the South American bird family tinamous, along with ratites—kiwis, ostriches, and the like—split off close to the base of the bird tree. Slightly later, chickens, ducks, and their kin branched away from the main group of birds. The subsequent history of birds has been enigmatic, but the new work offers some clarity. Songbirds, for example, are a sister group to parrots, and the two groups encompass all the descendents from their most recent common ancestor. Hummingbirds descended from night jars, evolving bright colors and a diurnal lifestyle along the way.

One of the more controversial results is that tinamous, all capable of flight, belong in the same group as the flightless ratites. This “can change the way people look at the evolution of flight,” Hackett says. Grouping the birds together suggests either that flightlessness evolved multiple times, not once in the ancestor to this group, or that flight evolved more than once in birds, showing up independently in the tinamous and in other flying birds. “This result flies in the face of many other kinds of data,” says Edwards.

## Shaky branches

The phylogenomics study of eukaryotes, conducted by Fabien Burki and Jan Pawlowski, both at the University of Geneva, Switzerland, and Kamran Shalchian-Tabrizi of the University of Oslo, Norway, also upsets old assumptions. Interested in deciphering the deep roots of eukaryotes, which include protists, plants, animals, and fungi, they combed the public databases, coming up with 135 genes from 65 species to compare. Based on the pattern of differences in the sequences, they and their colleagues came up with three early branches, two containing almost all eukaryotes and one tentatively placed branch representing excavates, protists that include Euglena and Giardia.

Unlike past analyses based on just a few eukaryotic genes, or just one, this phylogenomic effort, published online 3 June in Biology Letters, brought all photosynthetic organisms—save Euglena and its relatives—into one group. The researchers suggest that the cyanobacterium that gave rise to the modern chloroplasts seen in plants and in green and red algae was acquired much earlier in eukaryotic evolution than had been thought, though more data is needed to confirm this idea, says Burki.

That plants now group with dinoflagellates, diatoms, or freshwater flagellates—all previously considered independent “super-groups”—has raised some eyebrows. “I think this is untenable,” says Patrick Keeling at the University of British Columbia in Vancouver, Canada. Nonetheless, he adds, “this paper represents one of the right ways we should be going to resolve the tree of eukaryotes.” The challenge is to include more organisms in future studies. In doing so, “it's entirely possible that strong support for many relationships will evaporate,” he notes.

When Dunn and his colleagues wanted to tackle the animal kingdom, they couldn't find enough publicly available DNA sequence for the many species they needed to examine. So they sequenced 39.9 million bases from 29 of nature's more peculiar and little-studied creatures, including water bears, comb jellies, sea spiders, and a variety of worms. These data, combined with existing information, enabled them to evaluate 150 genes from 71 animals.

In some cases, the major branches of the new animal family tree confirmed researchers' suspicions. For example, based on a suite of similar traits seen in the animals, morphologists have long thought that mollusks all stem from a common ancestor. Yet there is no single unifying trait among the phylum, which includes scallops, squid, chitons, and snails. Many, but not all, have a toothlike structure called a radula, and a subset have no shell, even though mollusk means “thin-shelled.” Moreover, the molecular data did not back up the premise that all traditional mollusks belong together. Dunn's new tree shows that the mollusks are one big family, however. “It's nice to have tied [this relatedness] down,” says Dunn.

But the conclusion that comb jellies are the oldest animals is a surprise, says Dunn, who adds that the reaction has ranged from “‘That is so cool’ to ‘There is no way.’ “Dunn himself calls that result provisional and sees his 10 April Nature paper as just the beginning. Thanks to new sequencing technologies, “within a year or two, we'll be seeing studies that have 10 times as many genes from 10 times as many taxa,” he predicts.

And he's not the only one to soon be awash in data. Burki is generating more sequences for his work with eukaryotes, and Hackett and colleagues are expanding their data set as well. “Phylogenomics is becoming the rule,” says Hervé Philippe, who develops new phylogenetic techniques at the University of Montreal, Canada. Philippe looks forward to more phylogenomics studies that use gene order, even gene content or intron positions, to infer relationships—approaches that will become “more natural when complete genomes are available,” he says.

Philippe and others caution, however, that more data don't always guarantee better family trees. “It will be important to reanalyze [data sets] with many different and emerging methods to see if the results change at all,” says Edwards. And a few scientists question whether, even then, the full tree of life can really be resolved. But, Edwards argues, “phylogenomics is our best shot.”