News this Week

Science  21 May 2010:
Vol. 328, Issue 5981, pp. 958
1. Genomics

Synthetic Genome Brings New Life to Bacterium

1. Elizabeth Pennisi

For 15 years, J. Craig Venter has chased a dream: to build a genome from scratch and use it to make synthetic life. Now, he and his team at the J. Craig Venter Institute (JCVI) in Rockville, Maryland, and San Diego, California, say they have realized that dream. In this week's Science Express (www.sciencemag.org/cgi/content/abstract/science.1190719), they describe the stepwise creation of a bacterial chromosome and the successful transfer of it into a bacterium, where it replaced the native DNA. Powered by the synthetic genome, that microbial cell began replicating and making a new set of proteins.

This is “a defining moment in the history of biology and biotechnology,” says Mark Bedau, a philosopher at Reed College in Portland, Oregon, and editor of the scientific journal Artificial Life. “It represents an important technical milestone in the new field of synthetic genomics,” says yeast biologist Jef Boeke of Johns Hopkins University School of Medicine in Baltimore, Maryland.

The synthetic genome created by Venter's team is almost identical to that of a natural bacterium. It was achieved at great expense, an estimated $40 million, and effort, 20 people working for more than a decade. Despite this success, creating heavily customized genomes, such as ones that make fuels or pharmaceuticals, and getting them to “boot” up the same way in a cell is not yet a reality. “There are great challenges ahead before genetic engineers can mix, match, and fully design an organism's genome from scratch,” notes Paul Keim, a molecular geneticist at Northern Arizona University in Flagstaff. The “synthetic” bacteria unveiled this week have their origins in a project headed by Venter and JCVI colleagues Clyde Hutchison III and Hamilton Smith to determine the minimal instructions needed for microbial life and from there add genes that could turn a bacterium into a factory producing compounds useful for humankind. In 1995, a team led by the trio sequenced the 600,000-base chromosome of a bacterium called Mycoplasma genitalium, the smallest genome of a free-living organism. The microbe has about 500 genes, and researchers found they could delete 100 individual genes without ill effect (Science, 14 February 2003, p. 1006). But confirming the minimal genome suggested by those experiments required synthesizing a full bacterial chromosome and getting it to work in a recipient cell, two steps that have taken years because the technology to make and manipulate whole chromosomes did not exist. In 2007, Venter, Smith, Hutchison, and colleagues finally demonstrated that they could transplant natural chromosomes from one microbial species to another (Science, 3 August 2007, p. 632). By 2008, they showed that they could make an artificial chromosome that matched M. genitalium's but also contained “watermark” DNA sequences that would enable them to tell the synthetic genome from the natural one (Science, 29 February 2008, p. 1215). But combining those steps became bogged down, in part because M. genitalium grows so slowly that one experiment can take weeks to complete. The team decided to change microbes in midstream, sequencing the 1-million-base genome of the faster-growing M. mycoides and beginning to build a synthetic copy of its chromosome. Last year, they showed they could extract the M. mycoides natural chromosome, place it into yeast, modify the bacterial genome, and then transfer it to M. capricolum, a close microbial relative (Science, 21 August 2009, p. 928; 25 September 2009, p. 1693). The next step was to show that the synthetic copy of the bacterial DNA could be handled the same way. The researchers started building their synthetic chromosome by going DNA shopping. They bought from a company more than 1000 1080-base sequences that covered the whole M. mycoides genome; to facilitate their assembly in the correct order, the ends of each sequence had 80 bases that overlapped with its neighbors. So that the assembled genome would be recognizable as synthetic, four of the ordered DNA sequences contained strings of bases that, in code, spell out an e-mail address, the names of many of the people involved in the project, and a few famous quotations. Using yeast to assemble the synthetic DNA in stages, the researchers first stitched together 10,000-base sequences, then 100,000-base sequences, and finally the complete genome. However, when they initially put the synthetic genome into M. capricolum, nothing happened. Like computer programmers debugging faulty software, they systematically transplanted combinations of synthetic and natural DNA, finally homing in on a single-base mistake in the synthetic genome. The error delayed the project 3 months. After months of unsuccessfully transplanting these various genome combinations, the team's fortune changed about a month ago when the biologists found a blue colony of bacteria had rapidly grown on a lab plate over the weekend. (Blue showed the cells were using the new genome). Project leader Daniel Gibson sent Venter a text message declaring success. “I took my video camera in and filmed [the plate],” says Venter. They sequenced the DNA in this colony, confirming that the bacteria had the synthetic genome, and checked that the microbes were indeed making proteins characteristic of M. mycoides rather than M capricolum. The colony grew like a typical M. mycoides as well. “We clearly transformed one cell into another,” says Venter. “That's a pretty amazing accomplishment,” says Anthony Forster, a molecular biologist at Vanderbilt University in Nashville, Tennessee. Still, he and others emphasize that this work didn't create a truly synthetic life form, because the genome was put into an existing cell. At the moment, the techniques employed by Venter's team are too difficult to appeal to any potential bioterrorists, researchers stress. Nonetheless, “this experiment will certainly reconfigure the ethical imagination,” says Paul Rabinow, an anthropologist at the University of California, Berkeley, who studies synthetic biology. “Over the long term, the approach will be used to synthesize increasingly novel designed genomes,” says Kenneth Oye, a social scientist at the Massachusetts Institute of Technology in Cambridge. “Right now, we are shooting in the dark as to what the long-term benefits and long-term risks will be.” As ever more “artificial” life comes into reach, regulatory agencies will need to establish the proper regulations in a timely fashion, adds Oye. “The possibility of misuse unfortunately exists,” says Eckard Wimmer of Stony Brook University in New York state, who led a team that in 2002 created the first synthetic virus (Science, 9 August 2002, p. 1016). Venter says that JCVI has applied for several patents covering the work, assigning them to his company, Synthetic Genomics, which provided much of the funding for the project. A technology watchdog group, ETC Group in Ottawa, has argued that these actions could result in a monopoly on synthesized life (Science, 15 June 2007, p. 1557), but others are not worried. Given the current climate for granting and upholding patents of this type, says Oye, “it is unlikely that Synthetic Genomics will become the Microsoft of synthetic biology.” “One thing is sure,” Boeke says. “Interesting creatures will be bubbling out of the Venter Institute's labs.” 2. United Kingdom Will Britain's Coalition Wield the Funding Ax? 1. Daniel Clery The arrival last week of Britain's new coalition government, an unlikely union between the Conservative and Liberal Democrat parties, has brought mixed emotions for U.K. researchers. There has been a generally positive response to the choice of government ministers responsible for science. But after 13 years of a Labour administration that greatly improved the lot of scientists (Science, 18 May 2007, p. 965), and with the government deficit at record levels, there is grave concern that research funding will be hit. “The most important issue is to what extent cuts will fall on research and the best universities,” says astronomer Martin Rees, president of the Royal Society in London. As the Conservative–Liberal Democrat alliance took shape, researchers were pleasantly surprised to find David Willetts as the minister for universities and science. Although his background is in the humanities, Willetts was a Conservative spokesperson for education and science during the last Parliament. In a briefing earlier this week, Willetts said: “I understand the crucial importance of blue-skies research. Scientific research can't all be reduced to utilitarian calculations”—a very different message from that of earlier Conservative administrations. Robert Kirby-Harris, chief executive of the Institute of Physics, says he had “very positive discussions” with Willetts during Labour's reign. “We were impressed by him.” Willetts's boss at the Department for Business, Innovation and Skills, which oversees most science funding, is Liberal Democrat Vince Cable, who studied natural sciences and economics at the University of Cambridge. “They make a really strong pair of advocates” for science, says Hilary Leevers, acting director of the Campaing for Science and Engineering. Cable is a member of the cabinet and Willetts will attend cabinet meetings, although he is not a member. The most pressing concern for researchers is how their funding will fare, as the new coalition has made reducing the government deficit its top priority. The coalition has committed to cutting £6 billion from government spending this year, the broad details of which are due to be announced next week. But the specific impact on research spending is more likely to emerge from the government's “emergency budget,” due on 22 June, or a spending review this autumn that will outline funding for the next few years. The seven U.K. research councils, which distribute grants and manage research facilities, have been asked to come up with spending plans for a variety of funding scenarios, such as flat funding, a 10% cut, a 20% cut, and so on. “A 0% cut we could live with; 20% would be a total disaster,” says U.K. physicist Ian Halliday, president of the European Science Foundation. U.K. researchers point out that in response to the recession, some countries increased science spending as a way of boosting their economies. “Success breeds success,” says Rees. “The U.K. is strong in science, and it would be sad if anything happens to jeopardize that.” Willetts acknowledges that there are difficult times ahead. “It's going to be tough. … The boom has now come to an end.” 3. ScienceNow.org From Science's Online Daily News Site Golden Years Truly Are Golden It doesn't matter whether you're employed, whether your children still live at home, or even whether you're married. Life gets better after age 50. A new phone survey of hundreds of thousands of Americans confirms that people tend to be happier, less anxious, and less worried once they pass the half-century mark. The Scent That Makes Mice Run Scared Even if a mouse has never seen a cat before, he'll turn tail when one is nearby. Researchers suspected that the rodents somehow sniff out their would-be assassins, but exactly what they smelled was unclear. Now scientists have isolated the compound, one of a class of urinary proteins that are secreted by cats, snakes, and a variety of other predators. Vision Cells Help Set the Body's Clock If you like to surf the Web at night, beware: You may be affecting your body's internal clock. Humans, like other animals, rely on light cues to set their body's daily cycle of activity, or circadian rhythm. Now a new study shows that some wavelengths of light, such as those from computer screens, have an unexpectedly strong influence on these rhythms, keeping us awake, for example, when we should be sleeping. A Crack in the Mirror Neuron Hypothesis of Autism Brain cells thought to underlie our ability to understand one another work just fine in people with autism spectrum disorders (ASD), according to the authors of a controversial new study. Other researchers had proposed that these cells, called mirror neurons, malfunction in people with ASD, disrupting their ability to understand what someone else is experiencing. If the results hold up, researchers will need another way to explain the social deficits that characterize the disorder. Read the full postings, comments, and more at news.sciencemag.org/sciencenow. 4. Biomedical Research Former NIH Director Varmus to Head Cancer Institute 1. Jocelyn Kaiser Harold Varmus will be returning to Washington, D.C., to run the world's biggest cancer research program. President Barack Obama announced on 17 May that he plans to appoint the Nobel Prize–winning molecular biologist director of the$5.1 billion National Cancer Institute. Varmus, now president of Memorial Sloan-Kettering Cancer Center (MSKCC) in New York City, will replace current NCI Director John Niederhuber, a Bush appointee, who expects to remain at NCI as an intramural researcher.

The choice is not a surprise: Varmus's appointment has been rumored for weeks, although in early March he told Science that the idea was “far-fetched.” (He had announced in January that he would soon step down as MSKCC president.) But it is a surprise that Varmus, age 70, who directed the entire National Institutes of Health (NIH) from 1993 to 1999, would return to head the largest of its 27 institutes. Varmus, who was part of an informal search committee for NCI director that was having trouble finding an interested candidate, said the idea seemed “crazy” when a close friend suggested it, but he changed his mind: “Running the world's largest research program in cancer at a time when cancer research has never been more interesting seemed like a pretty good thing to do in the next several years.”

The most urgent task Varmus will face at NCI, however, will be to set priorities in an era of fl at budgets. For example, some researchers have questioned the value of “big biology” projects such as The Cancer Genome Atlas, a tumor mutation sequencing project that has cost $375 million so far. Varmus will also find on his desk an Institute of Medicine report calling for an overhaul of NCI's inefficient and underfunded cooperative groups, which have conducted many key cancer clinical trials. Varmus says he expects to review NCI's “entire portfolio.” NIH Director Francis Collins, who once worked under Varmus as head of the genome institute, praised the appointment. Collins said in a statement: “It is exhilarating and gratifying to have my good friend and colleague back at NIH. … Harold brings unmatched expertise at all levels.” Cancer researchers seemed thrilled. Varmus is “more qualified for this position than anyone else in the universe,” says cancer geneticist Bert Vogelstein of Johns Hopkins University in Baltimore, Maryland. “It's a spectacular coup on the part of the government,” said cancer biologist Robert Weinberg of the Massachusetts Institute of Technology in Cambridge. Edward Benz, director of the Dana-Farber Cancer Institute in Boston, says Varmus's “impeccable credentials” as a basic scientist, his time at NIH, and his decade of directing a cancer center give him “the ideal combination of experience.” Varmus shared the 1989 Nobel Prize in physiology or medicine with Michael Bishop for using a retrovirus to understand cancer-causing genes. During a tenure at NIH that included the start of a 5-year budget doubling, he cultivated the image of a rumpled academic who biked to work. At Sloan-Kettering, he oversaw an expansion of translational research programs and made key recruits. After endorsing Obama as a candidate in 2008 and advising the campaign, Varmus was named co-chair of the President's Council of Advisors on Science and Technology in 2009. Despite the praise for Varmus, some insiders are worried that his appointment signals an effort to curb the independence of NCI. Under the 1971 Cancer Act, the NCI director reports directly to the president and submits a draft budget directly to the White House. The bypass budget, as it's known, is mainly symbolic but still cherished by insiders. Varmus says he has no plans to push for congressional action to change NCI's special status. But “the reality is that NCI is part of NIH, and I plan to be part of the NIH.” A working group of advisers is currently examining NCI's special status, possibly to defend it. “Many people in the cancer community are concerned,” says medical oncologist Richard Schilsky of the University of Chicago Medical Center in Illinois, who chairs NCI's Board of Scientific Advisors. Varmus's return to government will mean a sharp pay cut for him from the$2.7 million he was reportedly earning at Sloan-Kettering to an NIH salary that usually tops out at about $300,000. The appointment does not require Senate confirmation. Varmus, who is bringing a small lab with him, expects to start at NCI on 12 July. 5. U.S. Science Policy House Blocks Bill to Boost Research Spending 1. Jeffrey Mervis Too much sex. Not enough money. That potent combination led the U.S. House of Representatives last week to reject legislation aimed at boosting research and education spending at three federal agencies. The 292-to-126 vote against the America COMPETES Act was fueled by concerns about the trillion-dollar federal deficit, a sentiment that could threaten the Administration's plans to double the research budgets of the three agencies over 10 years. But the lopsided defeat was also the result of a stealth attack by Republicans against a Democratic priority, using as a tool the National Science Foundation's handling of employees who trafficked in electronic pornography at work. Reauthorization of the 2007 COMPETES Act is a top goal for the year of retiring Representative Bart Gordon (D–TN), chair of the House Science and Technology Committee. The version that Gordon brought to the floor last week would have kept NSF, science programs at the Department of Energy (DOE), and the National Institute of Standards and Technology on a path for a 10-year budget doubling. It also added several new programs to foster innovation and bolstered the fledgling Advanced Research Projects Agency–Energy at DOE. Gordon had already pared almost$10 billion from an earlier version of the bill to accommodate Republican opponents, and he won bipartisan support within the committee for a 5-year, $82 billion bill. But Gordon was caught by surprise when ranking member Representative Ralph Hall (R–TX) instead proposed a 3-year freeze on spending at those agencies. Hall's amendment included language prohibiting payment of salaries to anyone who has been disciplined “for viewing, downloading, or exchanging pornography” on a government computer or at work. Although the amendment doesn't mention NSF, freshman Representative Lynn Jenkins (R–KS) cited a 2007 investigation by the NSF inspector general of employees who had engaged in such activity, flagging it as one of many agencies where the misuse of computers had occurred. Some 121 Democrats voted for the amendment to avoid being tagged as a friend of pornography before Gordon pulled the bill. Roughly a dozen people were found to have committed “serious infractions” involving pornographic material on their computers, says NSF Director Arden Bement, who estimates that “three or four” remain at the agency. “We nipped this in the bud,” says Bement, citing “aggressive steps” by the agency to clarify its policies and impose tighter filters. “But the law is kinda murky in this area,” he adds. Some of the cases “involved addiction, in which the people received counseling,” he notes, and several went to arbitration. He called pornography “a pernicious problem at many agencies” that requires “clear guidance from Congress.” At presstime, Gordon was hoping to get a second shot this week at passing the bill. But given the pressure to curb spending, the U.S. research community will need to work hard to persuade Congress that an increased investment is essential for the country's economic prosperity. And as an authorization bill, COMPETES at most offers guidelines for spending panels. “I hope that it's not an omen for what may happen with appropriations,” says Samuel Rankin of the Coalition for National Science Funding. 6. ScienceInsider From the Science Policy Blog A pretrial hearing before a federal judge in Tennessee was the most formal legal test yet for the use of lie-detection technology based on functional magnetic resonance imaging (fMRI) of brain activity. The hearing revealed new details about methods used in commercial fMRI lie-detection services; a ruling on whether to allow such evidence will establish an important precedent. The defendant in the case, accused of defrauding Medicare and Medicaid, hired a Massachusetts-based company that provides brain-scan lie-detection services to help establish that he is telling the truth. Biologist Dylan Evans was sanctioned for sexual harassment by the president of University College Cork in Ireland. Evans had showed a scientific paper that revealed that bats perform fellatio to several colleagues, one of whom protested. Evans says he was just sharing information, but the college's president said his behavior was “unacceptable.” The U.S. Supreme Court decided Monday that life sentences for most young criminals amount to cruel and unusual punishment. The decision in Graham v. Florida extended a 2005 ruling that banned the death penalty for juveniles. The American Medical Association, American Psychological Association, and other professional organizations submitted legal briefs that questioned whether youths should be held to the same standards of culpability as adults given research suggesting that their brains are still developing. Senators Joseph Lieberman (ID–CT) and John Kerry (D–MA) have introduced the American Power Act, which would require the country to reduce greenhouse gas emissions by 17% below 2005 levels by 2020. The House of Representatives passed an equivalent bill last year, but the challenge for Senate Majority Leader Harry Reid (D–NV) will be gaining the 60 votes needed to thwart a filibuster in an election year. It's unclear how much political clout the White House is prepared to exert to pass the legislation. See the full postings and more at news.sciencemag.org/scienceinsider. 7. Gulf Oil Disaster Five Questions on the Spill 1. Richard A. Kerr*, 2. Eli Kintisch*, 3. Lauren Schenkman* and 4. Erik Stokstad* Sizing up the weeks-long spill in the Gulf of Mexico and its impacts is proving a challenge for marine and coastal scientists. The source is beneath 1600 meters of seawater, the winds and currents spreading the oil can be capricious, and the marine life in the oil's path is spread over hundreds of square kilometers, from the sea floor to the surface. Commercial fisheries have already been affected, and fragile coastal marshlands are at risk. Just monitoring all these ecosystems is the first challenge; gauging the toll taken and sounding the all clear will come later. Here are five of the key questions that scientists will be trying to answer over the coming months and years. What's happening … to the oil? The magnitude of the catastrophe will depend on the oil's fate: the amount of oil released, how the oil is transformed chemically and physically, and how far and wide it travels. To date, scientists are far from answering any of these questions. The oft-cited 5000 barrels per day of oil spewing from the leaking well is almost certainly an underestimate. Scientists eyeballing videos of the sea-floor gusher or gauging the extent of the surface slick in satellite images see five or even 10 times as much oil coming out 3.5 weeks into the disaster. Continued analysis will improve those estimates, and ongoing efforts to stanch the flow may be informative, but as one oceanographer puts it, for now, “it is what it is.” Researchers have a better handle on how the oil is “weathering.” Samples collected from the sea surface show that, as expected, the oil is tending to lose its more volatile—and more toxic—components as it evolves from a simple liquid to an emulsified “mousse” to tarballs. Although well-aged tarballs are the least damaging form of lingering oil, those starting to appear on beaches are still so sticky that plants and animals could suffer greatly. Detergent-like dispersants applied offshore accelerate both physical and biological weathering, but chemists have yet to see obvious signs that dispersants are helping. Then there are biologists' concerns that dispersants could be affecting marine life in the open gulf directly through their toxicity or indirectly by causing more of the oil to linger far below the surface where fish and bottom-dwellers are. Researchers could soon get a better idea of what's happening to the oil and dispersants as well as where it's all going as field sampling gets in gear. Reports of large subsurface plumes of oil—perhaps enhanced by dispersants—are beginning to come in as sampling from boats and ships is extended to the subsurface. And two different autonomous underwater vehicles are scheduled to start mapping subsurface oil using optical sensors; one of them can return water samples for detailed analysis. The as-yet-loosely-coordinated effort to characterize the evolving spill is being conducted by the National Oceanic and Atmospheric Administration (NOAA) and the U.S. Geological Survey, while the National Science Foundation is supporting fieldwork through rapid-response grants. … to life on the sea floor? Two types of communities exist on the deep sea floor of the gulf. Where hydrocarbons seep out of the sediment, clams and mussels live with symbiotic bacteria that tap sulfide or methane for energy. In the same areas, polychaete tubeworms grow up to several meters long and can live for centuries. Elsewhere, corals capture prey that floats by or detritus that sinks from above. Expeditions have sampled sea-floor biota in the Gulf of Mexico for years with submersibles and ship-borne devices. In mid-May, a research vessel operated by the National Institute for Undersea Science and Technology (NIUST), a university consortium, began taking sediment and water samples from areas under and near the oil spill. Another group already had a passive sediment-sampling device in place nearby. When they retrieve the device, by October at the latest, they will know how much oil has settled onto the sea floor, either directly or entrained in seaweed. Most sea-floor studies to date have been funded by NOAA and the Minerals Management Service. NOAA is supporting the NIUST mission, which detected oil below the surface. The crew returned to port late last week and will begin analyzing samples. Meanwhile, they are beginning to plan another trip for follow-up sampling. … to coastal ecosystems? Coastal wetlands in the Gulf of Mexico have been under siege for decades. Chronic exposure to large amounts of oil could worsen their plight, killing marsh grasses and the creatures that live in the coastal sediment. Three weeks ago, academic researchers took sod samples from an established field site just east of the Mississippi River. Now they plan to collect oiled sods from the same site and, in a greenhouse lab, compare processes such as plant growth, photosynthesis, and soil respiration. In June or early July, the same team plans to survey sedimentation rates at 18 sites along the wetlands west of the Mississippi, an area likely to be hit by oil. A third study would assess the effects of fresh and weathered oil on different species of marsh plants; scientists say oil that has seeped into the soil and comes in contact with roots could have greater long-term impacts on vegetation than oil slicked on the surface. In the shallow waters of Louisiana's Breton Sound, where oil has already intruded, effects on marine life may already be visible. A team will collect live mollusks for tissue analysis, examine their shells for changes in growth rates, and look for deformities in the husks of foraminifera, an amoebalike bottom-dweller, and for large numbers of hibernating dinoflagellates in the soil. … to marine life? Just after the spill, researchers at the state-funded Dauphin Island Sea Lab off the Alabama coast stepped up their existing research to trawl for plankton along a 56-kilometer stretch south of the island; they will repeat the survey every 2 weeks. Another ongoing study based on rigs near the spill uses underwater, company-owned robots to monitor the squid, crustaceans, and fish that dwell 200 to 2400 meters below the surface. Besides tallying deaths, academic scientists will look for changes in the animals' daily feeding movements. An ambitious new study is about to start thanks to a rapid-response grant from the National Science Foundation. Working from two or three university research vessels, scientists from several institutions will trace oil over the next 3 to 6 months as it moves through the food chain from single-celled algae to large fish such as tuna. NOAA is monitoring the spill's effects on the more than 20 species of marine mammals, notably bottlenose dolphins and endangered sperm whales, and five species of endangered sea turtles that call the gulf home. Besides checking for oil in and on the bodies of six dolphins and more than 100 sea turtles collected so far, the agency is conducting aerial surveys to count the gulf's dolphins and whales, and taking biopsies of one bottle-nosed dolphin population to determine baseline levels of the animals' exposure to oil and other contaminants. Scientists will also monitor mammals acoustically with an underwater device. … to fisheries? On 18 May, NOAA shut down fisheries in a 118,000-square-kilometer area in the gulf. The move has threatened the lucrative shellfish industry. But the government says it is crucial to protect people from dangers of eating shellfish contaminated with polycyclic aromatic hydrocarbons, elements of oil that are carcinogenic. Scientists are scouring the area for tainted catch—so far, with the oil still offshore, none has been found—a tricky task in itself. Current analytical methods take days. So scientists at AOAC International, a nonprofit analytical chemistry group in Gaithersburg, Maryland, are working with testing companies to try to develop faster methods for preparing and analyzing samples with mass spectroscopy. A far more difficult task will be determining when it is safe to reopen the fisheries. After previous spills, NOAA has reopened fisheries when normal background levels of oil were detected in fish or shellfish samples. But given the size of the fishery affected and its critical importance to local livelihoods, such a strict standard may be unrealistic. Former Food and Drug Administration regulator David Acheson says the agency may have to develop new standards to certify fish that contains tiny amounts of oil above trace levels. But that could take “a very long time,” he says. “We don't really know what's safe.” 8. Biomedical Research Funding Battle of the Titans 1. Colin Macilwain* For the Howard Hughes Medical Institute and the Wellcome Trust, a rough decade on the stock market makes for straitened times. Loudon County, Virginia, and Somers Town, London, sit at opposite ends of the social spectrum. Loudon County is, in terms of median household income, the wealthiest of the 3141 counties in the United States. Somers Town is one of the poorest quarters of inner London, a working-class community hemmed in between the railway lines feeding into Kings Cross station. It is in Loudon County that the Howard Hughes Medical Institute has opened one of the largest freestanding biology laboratories in the United States, at Janelia Farm, which will eventually house about 500 people. In Somers Town, meanwhile, the Wellcome Trust is planning to help build the UK Centre for Medical Research and Innovation (UKCMRI), where 1500 scientists and support staff members will work cheek by jowl with London's people, hospitals, and universities. The different choice of neighborhood reflects the respective personas of the two largest biomedical research philanthropies in the world. Hughes is often seen as aloof and apart, and almost entirely dedicated to excellence in basic science. It keeps a safe distance from contentious public health and policy issues. Wellcome, in contrast, gets stuck right into almost every aspect of health research and policy. It supports the study of the history and sociology of medicine, gives grants to TV producers, seeks to influence policy, and collaborates deeply with the U.K. Medical Research Council (MRC) and other public bodies. Yet last November, Mark Walport, chief executive of the Wellcome Trust, signaled an important convergence in approach with that of Hughes. Wellcome, he said, would end its main grants for biology projects and programs in the United Kingdom—worth about £110 million annually—and instead support selected investigators for up to 7 years. The switch both excites and alarms British universities, where almost 3000 researchers currently receive Wellcome grants. “This is a massive change for us,” says Jonathan Weber, head of medical research at Imperial College London. “We relish the opening for our senior investigators. But we're anxious about the younger ones, since we lose the project grants that were the first step on the ladder for them.” Hughes's new president, University of California, Berkeley, molecular biologist Robert Tjian, meanwhile, is trying to coax a$500 million investment in Janelia Farm to bear fruit. The laboratory is a major departure for an organization that had always backed researchers inside existing universities or medical schools. The first 33 investigators are now bedding in at Janelia Farm, testing the bold prediction of its director, geneticist Gerald Rubin, that it can replicate the special environments of Bell Labs or the University of Cambridge Laboratory of Molecular Biology (LMB) in their heyday. Detractors see it as a gamble, however.

Financial pressures have curtailed the rapid expansion of both Hughes and Wellcome. Huge dips in asset values in 2001 and again in 2008 left both sitting on little more cash at the end of the decade than at its start (see p. 966).

At the same time, the two titans appear to be converging on a model in which excellent support for a small number of top university investigators, plus generous support for one or two intramural laboratories, consumes the bulk of their resources. Paul Nurse, the president of Rockefeller University in New York City and president-elect of the Royal Society in London, advises both philanthropies and applauds what he sees as Wellcome's shift in this direction. “Science in the U.K. has been held back by not having anything like the Hughes investigator program,” he says. “Hughes is now generating one or two Nobels every year; Wellcome hasn't been doing that.”

Billionaire's bequest

The Howard Hughes Medical Institute was created by the aviation billionaire back in 1953, mainly to provide a tax shelter for the Hughes Aircraft Co., whose entire stock it owned. Its research activities were modest until 8 years after Hughes's death, in 1984, when an imaginative Delaware judge, seeking to end a dispute over the estate, named eight prominent people as its trustees. They sold the corporation to General Motors the following year for a cool $5 billion, which became the institute's initial endowment. The structure of Hughes was novel: It was constituted as a research organization, not a foundation, and rather than offering grants or fellowships, it hired principal investigators inside the universities and medical schools as staff members. Under an agreement with the tax authorities, it was obliged to spend, on average, at least 3.5% of its worth annually through this mechanism and an additional$500 million over a 10-year period on related activities. Hughes trustees have chosen to continue these activities, such as grants and education, at about 1% of the endowment.

Wellcome has no such constraints but arrived at a similar spending rate—4.5% of the endowment—for a larger annual budget of £725 million ($1.1 billion) compared with$830 million for Hughes.

Donald Fredrickson, a former director of the U.S. National Institutes of Health (NIH), was appointed president by the new Hughes board and planned its move from Florida to a plush new headquarters in Chevy Chase, Maryland, near NIH, just north of Washington, D.C. But he left in 1987 after a scandal. (His wife was alleged to have meddled in Hughes management and to have created an unauthorized $200,000 account.) Fredrickson was succeeded by virologist Purnell Choppin, who hired Maxwell Cowan, a University of Oxford–educated South African neuroscientist, as scientific director. Choppin, who still has an off ice at Hughes, and Cowan, who died in 2002, stopped asking universities to appoint Hughes investigators, as they had done before, and instead selected them directly, with the help of topflight review panels. Cowan in particular set the tone: gentle-mannered, elitist, and almost ascetic, he believed that truly important science was only done by a talented few. “Max was enormously important and in many ways the key person in elevating the standard of quality at Hughes,” recalls former University of Chicago president Hanna Gray, the only surviving member of the original board of trustees and now its chair. Hughes's investigator model is a turbocharged version of the system used to promote scientists up the road at NIH, although Hughes has avoided entanglements with that agency. “NIH and Hughes were both doing their own thing,” says Choppin, “and it seemed logical to keep it that way.” With the arrival in 2000 of geneticist Thomas Cech as president, and Rubin as biomedical research vice president, Hughes gave even more autonomy to its investigators and broadened out into areas such as chemistry and computer science. “It became less formal and gave us more leeway,” recalls Bert Vogelstein, a cancer geneticist and Hughes investigator at Johns Hopkins University in Baltimore, Maryland. Hughes has sometimes been criticized as too cautious—simply backing the very best people, who would win NIH grants anyway. “They just sit there and skim the cream,” as one senior NIH official, who didn't want to be identified, puts it. But for those who are selected, there are immense advantages: reliable core funding and unprecedented freedom. “It's recognized around the world as the most effective way there is of supporting high-quality science,” says Nurse. More than one-third of Hughes's current investigators are members of the elite National Academy of Sciences, and 14 have won Nobel Prizes. An as-yet-to-be-published National Science Foundation–supported study by economists Pierre Azoulay and Gustavo Manso of the Massachusetts Institute of Technology and Joshua Zivin of Columbia University finds that very talented young researchers who become Hughes investigators “produce high-impact papers at a much higher rate” than those who don't. The downside is that they must undergo a bruising review every 5 years. “The system creates a huge amount of pressure and anxiety,” says one former Hughes investigator. At least the pain is brief: Investigators get a 45-minute presentation and 15 minutes of questions before a star-studded review panel decides whether to jettison them. Another doubt about the model concerns the generosity of the money, which sometimes needs to be spent in a hurry, especially at year's end. “The young people get the impression that money doesn't matter, and that's not a good lesson for them,” says the former Hughes scientist. Tjian says he's been aware of this problem and has changed the rules so that each investigator can carry over$200,000.

With four-fifths of its funding going to its 350 U.S. investigators, Hughes maintains a small headquarters staff of about 200 people. It has established an international program supporting, at different times, researchers in Eastern Europe, Latin America, and Canada. But this program is currently under review and likely to be reshaped to create closer ties to host institutions and more involvement in Asia. And it runs a \$40 million undergraduate education program, which has infused more research into science education at U.S. universities for 4000 students each year (Science, 16 April, p. 294).

The Hughes board's desire to do something that NIH could never do led to its 2000 decision to buy a site at Janelia Farm for a laboratory of its own (Science, 8 December 2006, p. 1530). “Whereas the Wellcome Trust is a major part of scientific infrastructure in Britain, we are just 2% of the NIH budget, and maybe 5% of basic biomedical research,” says Rubin, the laboratory's intellectual architect and now its director. “If we're going to have an impact, we need to do things that NIH is not doing.”

Timeline

See a parallel timeline of developments at HHMI and Wellcome Trust since 1935.

The Janelia Farm project fits that bill, Rubin says, by seeking to collect a group of unusually talented investigators and place them in a sublime research environment. “Our hypothesis was that there's something about the environment at Bell Labs and at the LMB at Cambridge that was special, that brought out the best in people,” he says.

Small research groups are fundamental to the mix, says Rubin, adding that successful investigators at universities often end up with so many people to supervise that they can't really do science themselves: “If you're going to be a scientist, working with your own hands,” he says, “then we're pretty much the only show in town.” The laboratory focuses on neuroscience, with a special emphasis on developing technologies to support neuroscientists.

Sean Eddy, a computational biologist who is working on algorithms to find homologous gene sequences, decided to move to the lab the day Rubin presented the idea to Hughes investigators. “The whole audience was hostile, but I thought: They're building my dream,” he recalls. “I just walked up to Gerry and said, ‘Sign me up!’”

Rubin has also attracted problem-solvers with nonacademic backgrounds such as Eugene Myers, the computer whiz who helped Celera produce a draft human genome; Eric Betzig, an acclaimed young physicist who dropped out of Bell Labs 10 years ago to work in his father's machine shop in Michigan; and Timothy Harris, a Bell Labs veteran now heading up an applied physics group at Janelia Farm.

Not everyone agrees that this decades-old model will work, however. According to Eddy, one LMB Cambridge veteran told Rubin that if he wanted to recreate that laboratory's original environment, he should build “a hut in a swamp.” And most Hughes investigators were strongly opposed to the idea of such a large, standalone lab—especially since the project won final approval after the 2001 market crash, when the investigators were being asked to absorb sharp cuts in their own budgets.

Rubin says the Hughes investigators are coming around slowly: “It's gone from 90:10 against to 50:50.” All say that it is too early to measure the success of a project that opened its doors in October 2006. Perhaps its strongest asset is Rubin's intense yet carefully reasoned pursuit of his ideal. “Our nightmare would be to recreate the Salk,” he says, referring to the prestigious but conventional Salk Institute in San Diego, California, “where it isn't clear that the whole is more than the sum of the parts.”

International reach

The bustling, palatial atrium of Wellcome's new headquarters on London's Euston Road houses a bigger staff than Hughes does—about 550 people—administering the trust's slightly larger and much wider range of programs.

The trust was set up through the will of the pharmaceuticals magnate Henry Wellcome in 1936, which handed it 100% ownership of his drug company. It has been active in medical research ever since, but its transformation into a scientific power-house only began in 1986, when the trustees began to sell the company in three tranches, leaving themselves with an endowment worth £6.8 billion in 1995.

“The Wellcome used to act like just another research council, playing second fiddle to the MRC,” recalls David Cooksey, a venture capitalist whose 2006 report to the British government has shaped national medical research policy, and who served on the Wellcome board of trustees from 1995 to 1999. “But they revisited the will, which says that the trust should seek to improve human health, and have become much more broadly spread.”

For U.K. biomedical researchers, Wellcome's mainstays are the 3- to 4-year project and program grants, which support the research activities of hundreds of university-based life scientists. The plan to replace these has come as a shock. “Once they move to the investigator awards, it is just going to be so competitive. If you've had the slightest dip in your career, you'll have no chance,” says a biologist at a top U.K. university. “It's going to be very, very tough for mature postdocs to get one of these.”

A 10-year strategic plan published in February unequivocally identifies the investigator awards as the first of three “focus areas” for the trust. Yet Walport denies that they mark a sharp change in direction. “This isn't a revolution,” he contends. “It is an extension of our existing fellowship program, building on our history of very successful funding of people.”

Wellcome has made the final calls for its closing grant schemes and announced on 5 May that the investigator awards will be worth between £100,000 and £425,000 a year, for up to 7 years. Walport won't say how many of them there will be: “People will need to make a case for how much money they need to answer their particular question,” he says, adding that an “early investigator” category will ensure that younger scientists, at the peak of their creativity, have a fair chance.

Wellcome has a far larger international program than Hughes does, spending about £100 million a year on activities aimed at typhoid, malaria, and other tropical diseases, mainly in Southeast Asia and east Africa. These stress clinical research and capacity-building. And last September, the charity announced a £90 million joint venture with Merck to build a center for the development of affordable vaccines in India. “The trust has had a long-term interest in tropical medicine, right from when Henry Wellcome had laboratories in Sudan,” explains James Whitworth, its head of international activities.

Wellcome is also more heavily involved in public outreach, spending about £40 million a year on activities ranging from the museum at headquarters—which attracted more than 300,000 visitors last year—to a regular, highly detailed survey of public attitudes toward science and medicine. In partnership with the government, it also runs a National Science Learning Centre at the University of York to do residential courses for schoolteachers: 6500 of them last year. “Our engagement program makes us different from other funders,” says Clare Matterson, the former management consultant who runs it. “I think we lead the way in this.”

Back in 1993, the trust established a major laboratory of its own, the Sanger Institute near Cambridge, to undertake the British arm of the Human Genome Project. Afterward, Sanger morphed into a large, multidisciplinary center whose 40 principal investigators do systems biology and genomics projects that take advantage of its sequencing and bioinformatics capacity.

Unlike Janelia Farm, Sanger has close ties with a nearby university—the University of Cambridge, where several of its investigators hold faculty positions—and many of its projects are collaborations with outside agencies and institutions. It gets £100 million a year from Wellcome and another £25 million from other funders. “Almost everything we do is collaborative,” says Michael Stratton, Sanger's director, whose own work is on the Cancer Genome Project. “It tends to be the sort of work that needs to be done on a scale that can't be achieved at other institutions.”

Wellcome's next major capital project is the UKCMRI, a £625 million collaboration with MRC, Cancer Research UK, and University College London, that will house up to 1500 scientists and support staff members. The laboratory won't have a hospital, but it will be focused on work that can be applied to human health: “We'll be trying to make sure that the status and reward systems are as tuned towards innovation as towards discovery,” says John Cooper, the project's interim chief executive, now running a small project team from Wellcome headquarters.

UKCMRI is illustrative of Wellcome's close ties with the U.K. government and of the risks and benefits that these carry. The government money isn't guaranteed, for example, given the perilous state of Britain's public finances.

Both Hughes and Wellcome have regained their financial footing after being roundly battered by the 2001 crash and then again by the one in 2008. The second blow was described in Wellcome's 2009 annual report as “the worst decade for equity markets in over 300 years.” Hughes's chief investment officer, Landis Zimmerman, says that the worst thing about the 2008 collapse was that it hit all investment sectors, “so that all that diversification doesn't protect us.” The financial performance of both institutions has been drastically less impressive since 2000.

Apart from one another, the two organizations have no rival outside government of equal size or with as much power to influence biomedical science. (The Bill and Melinda Gates Foundation, which spends more overall, devotes only a small fraction to scientific research.) Wellcome is “an excellent organization and its impact is huge,” says Leszek Borysiewicz, chief executive of MRC. Its existence allows Britain to spend almost twice as much on basic biomedical research as France or Germany. Hughes has set world-beating standards that other funders have struggled to match “I think this is the way all research should be funded,” says Vogelstein. “It encourages the best researchers just to do the best research. You are obliged to do something really phenomenal, not just incremental.”

Comparing the two organizations by measuring their relative scientific outputs is virtually impossible. Both eschew the current fashion for quantitative assessment, for example through citation analysis. David Pendlebury, a citation analyst for Thomson Reuters and one of the world's leading authorities on citation statistics, says that any statistical comparison will be “apples against oranges” partly because Wellcome investigators don't identify themselves as such when listing their affiliation. “What matters in science is what has been discovered, and who's been trained,” as Walport puts it, adding that narrative tends to describe these things better than statistics are able to.

In the end, Wellcome is larger than Hughes because it inherited more money. Hughes does more Nobel-league science, as it can cherry-pick talent from the world's only scientific superpower. Wellcome does more within its social orbit because it has chosen to do so, and because Britain is smaller. It also does more for the developing world, because of its imperial inheritance, and because Hughes's agreement with the tax authorities constricts its main research program to spending within the United States.

As pursuers of discovery by serendipity, it is appropriate that today's Hughes and Wellcome owe their very existences to serendipitous events. If either one had failed to liquidate its stock bequest at the right time, in the mid-1980s, biomedical research would be much the poorer. Perhaps the greed-is-good decade, so caricatured by novelist Tom Wolfe, had its good points, after all.

• Colin Macilwain is a writer in Edinburgh, U.K.

9. Evolution of Language

Animal Communication Helps Reveal Roots of Language

1. Michael Balter

An interdisciplinary gathering marks a turning point for a field historically richer in talk than data, but which now is increasingly embracing studies of animals.

UTRECHT, NETHERLANDS—Animals communicate with each other constantly: Birds sing, monkeys chatter, and apes pant-hoot. But what they say is usually pretty simple: They want to mate, send an alert about food or predators, or express their dominance in the group. Only humans appear to have true language, the ability to use abstract symbols—usually words—and combine them in a seemingly infinite variety of meanings about the past, present, and future.

Researchers have pondered the origins of language for at least 200 years, and for much of that time, their conjectures were little more than talk. Unlike many other human behaviors, such as art and toolmaking, language leaves no traces in the archaeological record. And many researchers have been doubtful about how much animal communication could reveal about the unique features of human communication. That began to change in the 1990s, when linguists, evolutionary biologists, psychologists, primatologists, and other scientists teamed up to test new hypotheses about how language arose (Science, 27 February 2004, p. 1316). Since 1996, this interdisciplinary crowd has gathered every 2 years at Evolang, a meeting devoted to deciphering the evolutionary origins of language.

Although some say the early Evolang gatherings suffered from too many hypotheses and too little testing, many think the meeting* here last month marks a turning point for the field. Participants flocked to hear a barrage of new data from animal and human studies. “The field has matured, and there is a trend towards more empirical work,” says evolutionary biologist W. Tecumseh Fitch of the University of Vienna in Austria.

One reason is that fewer scientists now follow the early views of linguist Noam Chomsky that language emerged de novo in humans, with little or no ape precursors. Indeed, Chomsky himself no longer holds strictly to that view, as evidenced by a seminal 2002 paper in Science he co-authored with Fitch and Harvard University psychologist Marc Hauser (Science, 22 November 2002, p. 1569), urging research into both the aspects of human language unique to humans and the aspects shared with other animals. “The more we study animals, the more we realize that they have abilities similar to ours,” says Natalie Uomini, an archaeologist at the University of Liverpool in the United Kingdom.

The new empiricism may help resolve one of the field's liveliest debates: whether the first human language consisted of gestures, similar to today's sign languages, or articulated speech. And here in Utrecht, a new and unlikely seeming animal model for human language got star billing: songbirds. Their ability to learn and imitate their parents' melodious tunes has many parallels with the ability of human children to learn spoken language, researchers say.

Hand, mouth, or both?

Pity poor Viki the chimpanzee. During the 1950s, two psychologists raised Viki in their own home like a human child and tried to teach her to speak. Viki managed a rough approximation of only four words: mama, papa, cup, and (maybe) up. The following decade, researchers had much better luck with a chimp named Washoe when they tried to teach him American Sign Language. But few scientists think Washoe's impressive efforts represent true language (Science, 2 April, p. 38).

Such evidence that apes are poor at vocalizing, but fairly good at gesturing, has bolstered the so-called gestural theory for language origins. According to this model, the first human language consisted of signing, and articulate speech came later. In recent years, the gestural theory has gained the upper hand in many scientific journals and meetings. “Apes are much better at controlling their hands” than at vocalizing, says Fitch. “Their gestures are more intentional and more under control.”

Many researchers have assumed that most primate vocalizations are innate or instinctual rather than learned, and so are uninformative about the origins of human language. For example, the vervet monkey gives out specific, stereotypical alarm calls corresponding to predators such as leopards, snakes, and eagles. These calls, which are innate, are a stark contrast to the way humans combine words in novel ways.

But psychologist Katie Slocombe of the University of York in the U.K. argues that the data don't support generalizing about all primate calls based on a few examples. In a poster, she and other European colleagues critiqued more than 550 studies of primate communication and found that few studies examined ape vocalization. “Absence of evidence may not reflect absence of [vocal] ability,” Slocombe's team concluded.

In other posters and talks, Slocombe and others documented that chimps in the wild do vary their vocalizations in response to circumstances, a step toward language. Slocombe and psychologist Klaus Zuberbühler of the University of St. Andrews in the U.K. showed that chimps modify screams they emit when under attack depending on the severity of the aggression and their social status compared with nearby chimps. Also, wild chimps emit so-called rough grunts—vocalizations associated with the finding of food—more often when chimp allies are present and the food is of high quality.

Such findings got dramatic support from a talk on a more distantly related species, the Campbell's monkeys of the Côte d'Ivoire. Primatologist Alban Lemasson of the University of Rennes 1 in France, Zuberbühler, and their colleagues found that the males of these forest-dwelling monkeys have six different types of calls, which the researchers refer to as Boom, Krak, Hok, Hok-oo, Krakoo, and Wak-oo. Yet these sounds are rarely used in isolation. Rather, they are combined in vocal sequences averaging 25 successive calls depending on whether the monkeys were encountering predators such as eagles or leopards, falling trees, the presence of neighboring groups, and so forth. Moreover, the animals carried on complex “conversations” in which the call sequences were constantly being modified or altered (see ScienceNOW, http://news.sciencemag.org/sciencenow/2009/12/04-02.html).

This complexity is “significantly beyond” what researchers have assumed for nonhuman primates, Lemasson told the meeting, and is “at odds with the gestural origins of language theory.” Uomini calls these studies “brilliant work” and says the call combinations could be considered a form of “proto-language” and “proto-speech.” And Erica Cartmill, a psychologist at the University of Chicago in Illinois, agrees that the Campbell's monkeys do seem to have “the ability to combine calls in different ways.” But she cautions that the calls fall far short of the kind of syntactical structures typical of human languages, which have specific rules for how words can be put together into sentences.

As the vocalization camp made gains, the gesturalists had advances of their own to put forth. Numerous recent studies have underscored the importance of gestures in both human and ape communication. In most humans, brain regions specialized for language, such as Broca's area, are located in the left hemisphere, which in right-handers also controls the movements of the right side of the body (see Science's Origins blog, http://tinyurl.com/n8wroy). Researchers are debating whether nonhuman apes also show asymmetries in homologous brain areas, and if these are the precursors of the lateralized language centers of the human brain.

Recent work by cognitive scientists Jacques Vauclair and Adrien Meguerditchian of the University of Provence in Aix-en- Provence concludes that such brain asymmetries in apes might indeed be linked to gesturing. The researchers found that baboons have a strong right-hand preference during communicative gestures such as begging for food but little hand preference during noncommunicative gestures such as wiping their faces. Captive chimpanzees show similar preferences, according to work reported by the team in Cortex this year.

And Vauclair has recently extended such studies to human children. Infants and toddlers tend to use their right hand for pointing—a communicative gesture that appears at about 11 months of age and closely accompanies early spoken language—even if they are ambidextrous or left-handed in other situations, Vauclair's group reported in Developmental Science last year. This suggests that human gesture and speech are linked and that both are at least partly localized in the brain's language areas, they concluded.

In Utrecht, Vauclair's Provence colleague Hélène Cochet presented further studies along these lines. She observed the pointing behavior of 48 toddlers in French day care centers. Earlier research has established two types of pointing behavior in young children: imperative pointing, which is used to ask for something the child wants; and declarative pointing, which is used to share interest or information. Researchers consider declarative pointing to reflect more complex cognitive processes, such as understanding that other people are independent agents with their own thoughts. On the other hand, most gesturing by nonhuman apes is only imperative, such as begging for food.

Cochet found that declarative pointing was more often accompanied by spoken utterances than was imperative pointing. And although children used their right hands more often for both imperative and declarative pointing than for noncommunicative gestures such as reaching for an object, the right-handed trend was even stronger when children were declaratively pointing to provide information to an adult. “Our results suggest that such cooperative gestures may have played an important role in the evolution of language,” Cochet told the meeting.

To explore whether ape gestures have specific meanings, Cartmill videotaped 28 orangutans at three European zoos, accumulating more than 100 hours of recordings. She identified 37 gesture types that could be reliably assigned to one of six different meanings, such as “play with me,” “share your food,” and “go away.” Thus the apes are conveying meaning to each other with their gestures, Cartmill concluded. “Intentional, meaningful, and socially sensitive communication emerged long before” the kind of symbolic communication typical of human language, she says.

But was that early human communication primarily gestural or vocal? Each camp continues to make its case, but some researchers at the meeting urged that the field acknowledge the importance of both. “Both modalities provide potential [primate] precursors for different elements of language,” says Slocombe, “but neither of them alone can provide the complete picture.” Thus primate vocalizations are discrete signals that can be combined in sequences—as in the Campbell's monkey—although primate gestures have the advantage of being flexible and highly intentional, Slocombe says. In his talk, Meguerditchian proposed that as early human language evolved, gestures might initially have been more effective for “talking,” although vocalizations might have been better suited for listening. Primate gestures appear more localized to brain areas homologous to Broca's area—implicated in speech production—whereas primate vocalizations have been more closely linked to brain areas homologous to Wernicke's area, which is involved in the understanding and perception of speech, he pointed out.

“Gesture … might have helped get speech off the ground over evolutionary time,” suggests psychologist Susan Goldin-Meadow of the University of Chicago. “The gesture-speech relationship we see today, where [they] work synergistically to form an integrated system, might have been there from the start.”

Birds move to center perch

Language evolution researchers have concentrated on apes and other primates because they are our closest relatives. But those animals can't match a key feature of human language: vocal learning, the amazing ability of young children to imitate the sounds of adults. Vocal learning does turn up in a handful of other species, including whales and possibly bats, but the masters of this talent are songbirds, parrots, and hummingbirds (Science, 31 January 2003, p. 646).

In his talk leading off a songbird workshop, biologist Johan Bolhuis of Utrecht University listed the numerous parallels between the way songbirds learn to sing and the way human infants learn to speak. Both must be exposed to adult “tutors”; juveniles of both species have a sensitive period for vocal learning; and both young birds and human infants “babble” (called “subsong” in birds) while learning to vocalize.

Over the past few years, Bolhuis and other researchers have traced vocal learning and song production to bird brain areas that appear analogous to human language areas such as Wernicke's and Broca's areas (see diagram). These similarities are not likely to be the result of shared evolutionary history, because the lineages leading to birds and humans diverged roughly 300 million years ago. But they may prove instructive all the same, Fitch says. “To do vocal learning, you need to … hear something and then pipe it over from the [brain's] auditory cortex to the motor cortex,” which controls speech production, Fitch argues. “There are probably not that many different ways of getting those connections.”

Thus, Fitch says, the bird model might be able to tell us “how to build a brain that can do vocal learning.” Researchers are beginning to find some of the molecular details of how that happens. At the Evolang meeting, Kazuo Okanoya, a biolinguist at the RIKEN Brain Science Institute in Wako City, Japan, reported that genes coding for molecules called cadherins—involved in nerve cell connections in humans and other mammals—are expressed at high levels when Bengalese finches listen to adult songs and down-regulated when they start to sing themselves.

Evidence for parallels between bird song and human language continues to accumulate. Some researchers have argued that only humans are able to distinguish words that closely resemble each other. But Verena Ohms of the Institute of Biology Leiden in the Netherlands taught zebra finches to distinguish two very similar-sounding Dutch words, pecking a button after hearing the correct word of either wit (white) or wet (law), even when the words were spoken by a variety of human voices, both male and female.

Fitch says that these parallels suggest that language evolution researchers can learn a lot about human speech by studying our distantly related feathered friends. He points to recent work by animal behaviorist Constance Scharff of the Free University of Berlin and her co-workers, showing that FOXP2, a gene implicated in human speech, also plays an important role in bird-song learning. Fitch says that such molecules might have been recruited by natural selection to perform similar functions even in species that went their evolutionary ways long ago. Thus, despite their distance from humans, birds are now perched firmly on the Evolang agenda. Indeed, the next meeting, in Kyoto, Japan, in 2012, will be organized by bird-brain expert Okanoya and his colleagues.

• * 8th International Conference on the Evolution of Language, Utrecht, Netherlands, 14–17 April 2010.