News this Week

Science  19 Apr 2013:
Vol. 340, Issue 6130, pp. 254

You are currently viewing the .

View Full Text

Log in to view the full text

Log in through your institution

Log in through your institution

  1. Around the World

    1 - Beijing
    More H7N9 Infections Reported
    2 - Washington, D.C.
    Supreme Court Weighs Human Gene Patenting
    3 - New York, New York
    Francis Crick Letter Sets Auction Record
    4 - Washington, D.C.
    Flawed Informed Consent Claimed in Preemie Study
    5 - Bad Berleburg, Germany
    Bison Back in Western Europe's Wild


    More H7N9 Infections Reported

    More flu.

    A girl infected with the H7N9 bird flu strain is treated in a Beijing hospital on 14 April.


    A proliferation of newly reported infections has scientists and public health officials increasingly concerned about the pandemic potential of the H7N9 avian influenza strain in China. As Science went to press, the World Health Organization (WHO) said it had received reports of 63 cases, 14 of them fatal, in six Chinese provinces and municipalities. Scientists haven't yet pinpointed the source of the virus or the route by which people become infected. Although there is no evidence that it is spreading efficiently between people, which would herald a pandemic, "the situation is very worrisome," says influenza researcher Masato Tashiro of the National Institute of Infectious Diseases in Tokyo. One of the infections is in a 4-year-old boy who had no symptoms; there could be many more people with no or very mild symptoms, Tashiro says.

    Last week, the China Center for Disease Control and Prevention in Beijing provided H7N9 samples to the other five laboratories in WHO's global network of collaborating centers for flu. These labs in turn can distribute the virus to other labs and companies that want to work with it.

    Washington, D.C.

    Supreme Court Weighs Human Gene Patenting

    On 15 April, the U.S. Supreme Court heard arguments in a case that presents a simple but radical question: Can human genes be patented? Decades after the court gave a green light to the patenting of recombinant DNA, a group of clinicians and researchers, led by the American Civil Liberties Union and the Public Patent Foundation of New York City, wants to invalidate patents on human breast cancer genes (BRCA1 and BRCA2) held by the diagnostic firm Myriad Genetics of Salt Lake City. The challengers say that DNA of this type is unpatentable because it is "a product of nature," not a human invention.

    A lower federal court accepted this argument in 2010, tossing out Myriad's patents. An appeals court last year backed Myriad. The challengers went to the Supreme Court. There, the justices probed both sides with questions and metaphors—Justice Sonia Sotomayor compared gene patenting to baking cookies—seemingly in search of a compromise that would retain patents on artificial DNA but not on "natural" genes. A decision is due by the end of June. Coincidentally, Australia's highest court was also preparing to hear arguments on the validity of Myriad's patents this week.

    New York, New York

    DNA diary.

    Francis Crick's letter to his 12-year-old son sold for a record $5.3 million.


    Francis Crick Letter Sets Auction Record

    "Jim Watson and I have probably made a most important discovery," begins a 1953 letter written by Francis Crick to his 12-year-old son. On 10 April, the 7-page letter, which describes the biologists' unraveling of the structure of DNA, was sold to an unidentified buyer for $5.3 million at Christie's, setting a record price for a letter.

    The next day, 11 other items from Crick's estate were auctioned off by Heritage Auctions, including the Nobel Prize medal that Crick won in 1962 (along with Watson and Maurice Wilkins) for the discovery of DNA's structure. The CEO of a regenerative medicine technology company bought the medal for a comparatively modest $2.3 million. Half of the proceeds of the Heritage Auctions sales will go to the Salk Institute for Biological Studies in San Diego, California, where Crick worked from 1976 to 2004. Another 20% of the proceeds will go to the London-based Francis Crick Institute, which is to open in 2015.

    Washington, D.C.

    Flawed Informed Consent Claimed in Preemie Study

    Researchers from 23 universities involved in a study of 1316 extremely premature infants failed to "disclose adequately" that the infants could become blind or die depending on the concentration of oxygen that they received, concluded a 7 March letter from the U.S. Office for Human Research Protection (OHRP). The letter was brought to a reporter's attention last week by Public Citizen, a Washington, D.C.–based nonprofit group. The infants were in the SUPPORT trial, which was led by the University of Alabama, Birmingham, and ran from 2005 to 2009. SUPPORT's results, published in 2010 in The New England Journal of Medicine, concluded that infants receiving lower oxygen levels (blood saturation target of 85% to 89%) were more likely to die, but less likely to become blind, than babies receiving higher levels (91% to 95%). Although both levels were within the accepted standard of care, Public Citizen suspects that "many, if not most, parents would have declined to enroll" in SUPPORT had they fully understood the risks.

    Bad Berleburg, Germany

    Bison Back in Western Europe's Wild


    European bison are back in the wild in Germany for the first time since the 18th century. Eight animals—one bull, five female adults, and two calves—were released on 11 April in a large private forest in the Rothaar Mountains, in North Rhine-Westphalia.

    After many decades of extinction in the wild, European bison populations have recovered thanks to captive breeding and reintroduction programs. Of the 4000 in the world, about 3000 are roaming freely in Eastern European countries including Poland and Belarus.

    But the German program is the first of its kind in Western Europe: a "historic moment," says Diana Pretzell, the World Wide Fund for Nature's manager for nature protection in Germany. The project benefits from the experience of previous reintroduction efforts, Pretzell adds: In particular, scientists provided advice to minimize inbreeding in the new herd.

    The reintroduction project was put forward by the forest's owner, logging businessman Prince Richard of Sayn-Wittgenstein-Berleburg. It has received a combined €1.58 million (about $2.1 million) from Germany's federal agency for nature conservation and North Rhine-Westphalia's state government.

  2. Random Samples

    Rolling With the Budget


    President Barack Obama's 2014 budget request to Congress demonstrates once again that, when it comes to science, "He gets it," says presidential science adviser John Holdren. Speaking last week at AAAS (which publishes Science), Holdren (at podium), offered an overview of the administration's proposed $143 billion investment in research and development. He was joined by, from left: White House official Patricia Falcone, NASA Administrator Charles Bolden, National Institutes of Health Director Francis Collins, acting National Science Foundation Director Cora Marrett, and acting National Oceanic and Atmospheric Administrator Kathy Sullivan. The 2014 request is $12 billion over the sequestration-depressed amount allocated this year and represents a 1.3% boost over 2012 levels. Civilian research would grow by 9%, while defense spending would shrink by 5%. For full budget coverage, see page 257.

    By the Numbers

    $3000 to $5000—Cost to generate a human genome sequence in 2013, compared with $1 billion when the Human Genome Project started in 1990 and $10 million to $50 million when it ended in 2003.

    40 to 55—Percent of Tamiflu prescriptions that went unused during the 2009 to 2010 H1N1 swine flu pandemic, according to a study in PLOS ONE that estimated drug compliance rates based on sewage water.

    $13.6 billion—Amount that Thermo Fisher Scientific of Waltham, Massachusetts, will pay for Carlsbad, California–based biotech equipment manufacturer Life Technologies.

    Long-Lost Mars Lander Found?


    About a week after the unmanned Soviet probe Mars 2 crashed into the planet in 1971, its twin, Mars 3, became the first spacecraft to manage a soft landing on Mars. Mars 3 looked around for about 15 seconds before its transmission mysteriously stopped. On 11 April, NASA announced that its Mars Reconnaissance Orbiter may have spotted the errant Mars 3 lander, along with detritus from its descent—its parachute, heat shield, and terminal retrorocket. The images, taken with the orbiter's High Resolution Imaging Science Experiment (HiRISE) camera, were actually captured about 5 years ago; an online community of Russian space enthusiasts pored over the images to find viable candidates for the expected hardware. The features appear to be a good match to the lander's components, HiRISE Principal Investigator Alfred McEwen said in a NASA press release, but further analysis "may help to confirm this interpretation."

    Science Live

    Join us on Thursday, 25 April, at 3 p.m. EDT for a live chat on the Anthropocene. Have humans altered the planet enough to earn our own epoch? When did it begin?

  3. Newsmakers

    IVF Pioneer Dies

    British physiologist Robert Edwards died last week in his sleep at age 87, nearly 35 years after he helped usher the first "test-tube" baby into the world. Edwards and gynecologist Patrick Steptoe developed the technique of in vitro fertilization (IVF), which has helped millions of couples around the world have children, and for which Edwards was awarded the Nobel Prize in physiology or medicine in 2010. (Steptoe died in 1988.)



    Edwards's efforts began in the 1950s, when he studied how human eggs mature and under what conditions fertilization can succeed. He reached out to physician Steptoe to help bring these lab discoveries to couples battling infertility. The two transferred dozens of embryos before one developed into a healthy pregnancy: Louise Brown was born in 1978.

    "Probably not since the invention of nuclear weapons has a scientific advance been received with such mixed feelings," wrote The New York Times in an editorial published 3 days later. While cutting-edge reproductive technologies still ignite debate, the ethical handwringing over IVF has long since settled down, and the treatment has become a mainstream part of fertility care.

  4. Trachea Transplants Test the Limits

    1. Gretchen Vogel

    More than a dozen ill people have received a bioengineered trachea seeded with stem cells during the past 5 years, but outcomes are mixed, and critics say the treatment may not do what its developers claim.

    Pushing forward.

    Paolo Macchiarini has treated more than a dozen patients with bioengineered tracheas.


    Five years ago, Claudia Castillo was struggling to breathe. A severe case of tuberculosis had damaged the 30-year-old Barcelona mother's airway so much that she couldn't walk up stairs, much less keep up with her two young children. Out of other options, she agreed to help make medical history when her surgeon offered her a radical proposal. In June 2008, she became the first patient to receive a trachea transplant that made use of her own stem cells—the most advanced example at the time of tissue engineering.

    The operation, she told various media in the following months, gave her back a normal life. It also made medical celebrities of the surgeons who developed and implanted the artificial trachea: Paolo Macchiarini, Martin Birchall, and their colleagues. They were hailed as pioneers leading the world toward an amazing future of regenerative medicine in which doctors will make replacement parts to order. Since then, 14 other patients have received bioengineered tracheas.

    However, almost all the cases so far have been done under "compassionate use" rules, in which untested procedures are allowed in seriously ill patients for whom conventional treatments are unlikely to help. That makes the outcomes more difficult to evaluate than if the procedures were performed as part of clinical trials. Indeed, some observers wonder if the clinical feats are really examples of successful tissue engineering that utilize the special abilities of stem cells—or an elaborate temporary fix that is destined to fail. More worrisome is that at least two recipients of these bioengineered tracheas have died under circumstances that that might be related to their transplants.

    "The motivation is more than noble," says Harald Ott, a surgeon at the Massachusetts General Hospital Center for Regenerative Medicine in Boston. "The surgeons are trying to help the patients." However, data from animal studies using the technique are sparse, he and others note. "If you translate something [to patients] very quickly, you don't have the luxury of getting the readout of what's really happening. … That's why we do animal studies, so that we can see what happens inside the grafts." In fact, one critic says it is impossible that the trachea implants are working as their makers propose, with new cells taking over the structure and creating living tissue. "The claims are hot air," says Pierre Delaere, a head and neck surgeon at the University Hospital Leuven in Belgium who has written two letters to The Lancet criticizing the work.

    Both Macchiarini and Birchall are trying to address the doubts about their approach. Now working independently, each has a major grant to conduct traditional clinical trials of stem cell-laden trachea scaffolds, one in Russia and another in the United Kingdom. And Macchiarini, who has won more than $10 million in grants from the European Union and Russia, has received U.S. Food and Drug Administration (FDA) approval to conduct the first such transplant in the United States.

    But Delaere, Ott, and others contend that, before pushing ahead with the clinical trials, the trachea pioneers should focus on understanding what happens to the grafts over the long term. Failing to do so could endanger patients and ultimately damage the field, they say. Although these initial operations have been an important step forward, "they've got a lot of work to do to make a procedure that is available widely to patients," says Alan Trounson, a stem cell researcher and president of the California Institute for Regenerative Medicine in San Francisco. "In the end, do we have tissue that's going to be functional for the lifetime of the person? I'm unsure whether we really do."

    A delicate balance

    Although at first glance the trachea might seem like a simple tube, its thin but cartilage-reinforced walls must stand up to near-constant use as a person breathes, clears his throat, or coughs. Any transplant, therefore, has to be strong enough to withstand such pressures without collapsing. But a rigid prosthesis can rub against and damage the adjacent major blood vessels in the upper part of the chest, leaving a patient at risk for a fatal hemorrhage. At the same time, the natural blood supply for the trachea's tissues is intricate, with vessels too small for surgeons to easily reconnect during a transplant operation. And because it is exposed to inhaled air, the wound between the implant and the remaining airway is especially vulnerable to infection.

    Surgeons have tried for years to find ways around these challenges, without much success. When Castillo was hospitalized in Barcelona in March 2008, Macchiarini, who was then at the University of Barcelona's Hospital Clínic, and Birchall, then at the University of Bristol in the United Kingdom, had experimented with bioengineered transplants in pigs. They would take a trachea from a pig and remove its living cells to create a so-called decellularized scaffold. They seeded this with cells from the recipient pig: bone marrow cells on the outer layer, thought to help form new cartilage, and epithelial cells on the inside, which they hoped would regrow the trachea's lining. They allowed the cells to grow on the scaffold for several days in a bioreactor designed to provide different conditions for the two types of cells. They hoped that the decellularized scaffold would not require immunosuppressive drugs to prevent its rejection and that the seeded cells would take over the removed cells' roles, ultimately forming a living organ.

    Transplant success?

    Claudia Castillo, the first recipient of a bioengineered trachea, with her children several months after the 2008 operation.


    The pig studies were still unpublished, but the researchers had observed that the animals could survive for 60 days with the implants. Castillo, who was suffering from recurring infections, was facing the possibility of losing one of her lungs. Macchiarini proposed that instead they try to replace part of her trachea using the pig-tested technique, and she agreed.

    Birchall readily admits that when the operation took place, "we were a million miles from what we thought we'd need to know before trying it in people." Regulators in both Spain and the United Kingdom, where Birchall's group seeded Castillo's cells on a decellularized trachea from a deceased donor, eventually allowed the procedure without the kind of preclinical testing usually required. "We ran roughshod over regulations—with permission," Birchall recalls. "It wasn't done to the highest possible standards. But it worked, and it changed the way people think about regenerative medicine."

    The medical team was aware of the risks, he says. "We had the naïve confidence of youth, I suppose. We believed that it would work. Paolo is a very courageous and skilled surgeon. His was the real courage. My job was to brave the regulatory environment and fit the science to what we needed."

    The surgery went smoothly and was soon featured in news stories around the world. As the researchers reported in The Lancet in December 2008, Castillo was released from the hospital without complications after 10 days; several months later, she could again walk several flights of stairs, care for her children, and had normal lung function.

    Complications arose a few months later, however, when Castillo's airway narrowed again and collapsed, requiring doctors to insert a stent—a semiflexible tube—to keep her airway open. They removed it 6 months later, Macchiarini says, but she has had to have several more inserted and removed over the past 5 years. He says she is now fine and has had her latest stent removed. "Thanks to regenerative medicine, she still has her both lungs and enjoys a superb quality of life," Macchiarini says. "Thanks to her we have learned so much and opened the door for more advanced applications." (Despite multiple attempts to reach Castillo, she did not comment for this story.)

    Macchiarini, who moved to the Karolinska Institute in Stockholm in 2010, says that he has treated eight more patients with decellularized and reseeded tracheas, three of them children. In six other patients, he has used a different technique, seeding cells on a synthetic scaffold made of a nanocomposite polymer designed to mimic cartilage instead of a donor trachea.

    Not all of the stories have happy endings. One of the children, Birchall says, died of the cancer that necessitated the operation in the first place, although "it gave her quality of life for the last few months." Macchiarini says another adult patient also died of the cancer that had damaged the trachea.

    At least two other patients have died suddenly. One was a child who had massive bleeding into her chest after receiving one of the decellularized, reseeded tracheas; "we don't really know what happened," Birchall says. The other, Chris Lyles of Baltimore, Maryland, died on 5 March 2012 after receiving a synthetic scaffold seeded with his stem cells in November 2011 in Stockholm. Lyles had returned to Baltimore in January and was apparently doing well initially. The family did not make his cause of death public, and according to Macchiarini, there was no autopsy, although he suspects pneumonia based on the medical report.

    But thoracic surgeon Alain Wurtz of the Lille University Teaching Hospital in France says both cases could also be consistent with the rupture of major blood vessel, which the implant might have caused. (Wurtz has treated several patients who need new tracheas by transplanting part of an aorta from a tissue bank into the neck.)

    The researchers have published two case reports in addition to Castillo's. In July 2012, Birchall and his colleagues described in The Lancet the 2-year follow-up of the first child to receive a transplant. The paper notes that the boy, whose own trachea had never developed properly, needed several stents in the months following the transplant, but by 18 months his chest CT scan and lung function were normal and he had returned to school.

    In 2011 in The Lancet, Macchiarini and his colleagues described the 5-month follow-up of Andemariam Beyene, a geophysics graduate student in Iceland who had a large tumor growing on his trachea. He received the first synthetic scaffold. His current doctor, Tomas Gudbjartsson of Landspitali University Hospital in Reykjavik, tells Science that Beyene has had several stents, but is healthy enough that he was able to complete his studies last year. The researchers have mentioned other patients in passing in several papers, but no formal reports have been published about their health, and Science has not been able to independently verify the current status of all the patients.

    Lingering questions

    What's been published and presented at meetings so far hasn't satisfied critics of the stem cell approach. "They should not tell the world that they did the first synthetic organ transplantation," Delaere says. Simply wrapping an artificial stent with tissue called omentum is known to be a temporary treatment for patients who need new tracheas and that, in essence, is what has happened in the recipients of the cell-seeded tracheas, Delaere contends. (Macchiarini and his colleagues wrap the transplants in omentum to reduce inflammation and promote blood vessel growth.) "If they claim something miraculous, they have to show corresponding data. They don't do that," says Delaere, who has developed his own solution to damaged tracheas: taking an intact trachea from an organ donor and implanting it in the patient's forearm for a few months, allowing new blood vessels to grow before it is transplanted into the neck. The Lancet reports, he charges, do not provide convincing evidence that the graft is functioning properly. Wurtz agrees: "There is absolutely no detailed data about these patients," he says.

    Macchiarini and Birchall, who is now at University College London, acknowledge that it is difficult to tell what exactly has happened inside the implanted tracheas. They say, though, that clinical outcomes are their primary concern and the current stable health of more than half of the recipients is strong enough evidence that the technique works. Both are now working on clinical trials of the technique, although they are no longer working together, in part because of a dispute over the contents and authorship of one of the transplant papers.

    In March, Birchall received a £2.8 million ($4.3 million) grant from the United Kingdom's Medical Research Council to conduct a trial of decellularized and stem cell–seeded upper trachea and larynx, with roughly 10 patients. Macchiarini has already completed two transplants in Russia as part of a clinical trial—funded with a $6 million grant from the Russian government—that he says should eventually enroll 20 or 25 patients. "We were allowed to do this type of transplantation only in extreme cases," he says. "The clinical study for the first time gives us a chance to include patients who are not in such critical shape."

    Macchiarini is also the lead investigator on a 5-year, €4 million ($5.2 million) grant from the European Union to begin a clinical trial using decellularized tracheas and further develop the polymer scaffolds in large animal models. That project may need to be reorganized, however, following a legal dispute last year in Italy, where the transplants were supposed to take place—Macchiarini had a part-time position at Careggi Hospital in Florence. In September, however, Italy's financial police accused him of attempted extortion, and briefly placed him under house arrest, for allegedly telling a patient that he could receive treatment in Germany for €150,000. Macchiarini and his lawyer say that he was simply informing the patient of possible options, not demanding payment. The main charges were soon dropped, but Macchiarini says that the charges stemmed from academic politics in Tuscany and he has severed ties with the hospital and university there. "There is no way to go back there."

    Not so simple.

    The trachea extends from the larynx to the chest and is formed from a thin membrane supported by cartilage rings and sustained by blood vessels.


    A solid base?

    The questions that remain 5 years after Castillo's transplant are highly relevant for the field of tissue engineering as a whole. When other researchers have used cell-seeded scaffolds to repair blood vessels, they have shown that it is not the seeded cells that ultimately populate the graft. Rather, the seeded cells produce signals that both dampen the body's inflammation response to the transplant and attract other cells from surrounding tissue. These immigrants ultimately take up residence and form the new tissue (Science, 26 August 2011, p. 1088).

    No one knows whether something similar is happening in these engineered tracheas. The detailed analysis of the vascular implants was conducted in mice because it's relatively easy to genetically tag mouse cells and trace where they end up. Mouse tracheas, however, are extremely small, making surgery impractical. Most of the animal studies on trachea replacement have been in pigs, sheep, and rabbits, and most researchers have not attempted to find out exactly how the replacement trachea is repopulated.

    That is a missed opportunity, says Christopher Breuer, a pediatric surgeon at Nationwide Children's Hospital in Columbus. Breuer has used cell-seeded grafts to repair severe heart defects in children and led the work in mice that demonstrated the details of how the grafts work. "I think it is important to take a step back and look at what is happening" in animal models before the clinical trials start, he says. "Ultimately, the half a step back will allow you to take many steps forward." Macchiarini has done additional lab-based experiments to find out what happens to decellularized tracheas over time. After a year in saline solution, he and his colleagues found, human tracheas became significantly weaker and lost some of the microstructure that is thought to help new cells take up residence and regenerate healthy tissue. "It's not surprising. That's physics in action," says Doris Taylor of the Texas Heart Institute in Houston, a co-author on the study, which was published in Biomaterials last year. In the body, incoming cells probably mitigate the effect, she says, but the details are not clear.

    Those results, and the collapsing airways of Castillo and other initial transplant recipients, were what prompted Macchiarini to switch to the synthetic scaffold, developed by materials scientist Alexander Seifalian at University College London and other researchers. The Russian clinical trial is using the lab-made scaffolds, and this month Macchiarini is scheduled to implant one in a child for the first time. The patient is a 2-year-old born with an undeveloped trachea. FDA has given its approval under its Investigational New Drug rules, which allow for exceptions to full approval.

    But even controlled clinical trials won't reveal as much as well-designed animal studies, Ott says. That means that it is difficult for other groups to build on what has been published, he says. Taylor concurs. "We are never going to understand everything," she says. "But we have to understand enough to make it reproducible, safe, and effective."

  5. Profile: Laurent Keller

    Chasing Ants—and Robots—to Understand How Societies Evolve

    1. Elizabeth Pennisi

    Laurent Keller's passions go far beyond ants as he taps genomics, robots, and other approaches to answer evolutionary questions.

    LAUSANNE, SWITZERLAND—Kenneth Ross still vividly recalls the day that Laurent Keller saved his life. Ross, an entomologist at the University of Georgia, had been so focused on looking for fire ants in rural Georgia that he stepped into the road just as a logging truck roared by. Fortunately, Keller was quick enough to grab Ross out of the truck's path. "I can still smell the pine rushing past," Ross says of the 1993 incident.

    Social scientist.

    Laurent Keller is serious about his research on social behavior in insects, but colleagues say he also is fun to be with.


    Keller's quick reflexes and decisive actions have held him in good stead in his science as well. At the time, Keller was a young Swiss postdoc with poor English skills but a lot of enthusiasm. Twenty years later, at the University of Lausanne, he has more than 280 publications that testify not only to his language proficiency but also to his ability to jump quickly into emerging fields and make early critical insights. "He always has a good feel for where a field is heading and is always at the forefront," says Jürgen Liebig, a behavioral ecologist at Arizona State University, Tempe. That, coupled with persistent attention to key problems, has led Keller, 52, to success on multiple levels, including receiving the E. O. Wilson Naturalist Award in 2005 from the American Society of Naturalists.

    Almost 15 years ago, Ross and Keller came up with one of the first genes underlying social behavior. Just a few months ago, Keller's team reanalyzed that "gene" and showed that it is actually a large segment of inverted DNA that encompasses 600 genes. In the meantime, Keller delved into social conflict, crafting experiments that shore up key theoretical ideas about how societies work. He also teamed with engineers and computer programmers to build robots that "evolved" social behavior. That work "planted the seeds for a whole new approach to understanding social evolution," says Gene Robinson, an entomologist at the University of Illinois, Urbana-Champaign. Most recently, Keller's lab has shifted into studies of collective behavior in ants, using a newly developed tracking system that can follow up to 225 individual ants 24/7 for weeks at a time (see sidebar, p. 270). "We try to do stuff that [other] people don't," Keller says.

    Keller "has a knack of picking important questions, picking the right system, and having the persistence in chasing them," says Andrew Bourke, an evolutionary biologist at the University of East Anglia in the United Kingdom. And his work has relevance far beyond ants. "Understanding how individual behavior is specified in large groups [can lead] to principles of social organization that may explain the organization of other groups, including humans," says University of Lausanne neurobiologist Richard Benton. In ants, unlike humans, "you can test your hypotheses about how communication between individuals is going to affect the group as a whole," Benton adds.


    As an undergraduate at the University of Lausanne, Keller knew he wanted to study social behavior in animals. Great apes were his first choice, but he realized that there would be a low return per hour of fieldwork. Zoo studies present an artificial environment. So he settled for ants.

    He became fascinated with fire ants, South American invaders that have spread rapidly across the southern United States in the past 80 years. In particular, he wanted to understand why some fire ant colonies support a lone queen, while others form supercolonies with multiple queens.

    He struck up a close collaboration with Georgia's Ross, at one point serving a joint postdoctoral fellowship with Ross and Harvard University's famed ant researcher Edward O. Wilson. Several times during the late 1980s and 1990s, Ross and Keller traveled to Latin America, spending weeks on the road crisscrossing thousands of kilometers in search of native populations of fire ants. More than once, Keller, in his faulty Spanish, negotiated their way out of sticky situations—gauchos riding up with guns drawn, or border police trying to shake them down for bribes. He explained to all comers that they were legitimate researchers and must continue their work—and he got his way. "He's like a bulldog, he doesn't give up," Ross says. "He does the same thing with his science—he's tenacious."

    One example of that tenacity is his and Ross's 15-year quest to understand the genetic basis underlying the multiqueen fire ant colonies. In 1997, lacking the technology to look for DNA differences, Ross had fingered a protein called Gp-9 that seemed to differ in ants from the two types of colonies. In 1998, he and Keller showed that the still unidentified gene for Gp-9 had both a dominant and a recessive version, or allele. Workers with both versions of the allele accepted multiple queens, provided those queens also had both alleles. Otherwise the workers killed the extra queen.

    Gp-9 is an odor molecule found all over an ant's surface. Yet, the different versions of the gene seemed to also determine worker behavior, the colony's social organization, and even the size of the queen. "It started to become difficult to think that it was a single gene for all of those traits," Keller recalls. But at the time, Ross and Keller didn't have the technology to explore further.

    Still, they kept the problem in mind. When the revolution in DNA sequencing put the genomes of many organisms within easy reach, Keller acted. In 2008, he helped organize the effort to decipher the genome of the fire ant Solenopsis invicta. Published in 2011, the genome paved the way for Keller to find out about Gp-9. Sequencing doesn't physically place genes on chromosomes, so Keller used a technique called RAD-sequencing (Science, 25 February 2011, p. 1006), which generates thousands of landmarks on chromosomes, to pinpoint Gp-9. Thus, he could see which genes were linked to it, as those linked genes might also be involved in prompting tolerance of multiple queens.

    To his team's surprise, they found a 13-million-base block around the Gp-9 gene that was passed down to offspring as a single unit—essentially a "supergene." During reproduction, matching chromosomes from the male and female usually cross over and recombine to create new mixtures of alleles, but this doesn't happen with this stretch of DNA. As the team reported in the 31 January issue of Nature, the block, which contains more than 600 genes, is inverted with respect to the DNA on the matching chromosome, thereby preventing recombination. Thus, each parent's contribution remains unchanged as it passes on to the next generation. "I was surprised it was so big," says Keller, who dubbed the DNA a "social chromosome."

    The discovery "opens an entirely new area of research for this system," says Mathieu Joron, an evolutionary geneticist at the National Museum of Natural History in Paris. One challenge now is to determine which of those 600 genes set the stage for multiqueen colonies.

    Against the grain

    But Keller likes challenges. Although he says that he doesn't take himself too seriously, beneath his broad smile and welcoming eyes is an intense, strongly opinionated man who doesn't hesitate to speak his mind or criticize those around him. "He never seems to get angry, but will state very plainly and in a nonemotional way, 'This is worthless,' " says his postdoctoral fellow Eric Lucas. Keller signs his reviews, so colleagues know whose blunt words helped make or break a paper. "Some people are rubbed the wrong way," Ross says.

    Several lines of inquiry in his lab have gone against the grain, including work on how insect caste is determined. "He's really been looking at a variety of different species that get us thinking outside the box about ways to maintain different castes," explains Andrew Suarez, an ant ecologist at the University of Illinois, Urbana-Champaign.

    Communication 101.

    In this evolution experiment, robots (glowing blue, above) move (inset) to find the "food" and signal other robots.


    Researchers had long assumed that environmental factors determined whether a female social insect became a queen or a worker, as both types have the same set of genes. That's how it works in honey bees. But in the mid-2000s, several exceptions emerged in ants. In each case, Keller's group or others noticed unexpected genetic differences, for example between queens and males. With such strange data, "most people would have given up," says evolutionary biologist Daniel Kronauer of Rockefeller University in New York City. "[Keller] noticed what was going on and what the implications would be."

    Through a series of experiments, Keller's team showed that these ants have an unusual reproductive biology. In one case, for example, workers were sexually produced—a combination of male and queen genetic material. But queens were parthenogenic and were clones of each other. And the queen's DNA was discarded in male offspring, making all males clones of each other. "It's as if you had two species' genomes that only come together in the sterile worker," Keller explains.

    In these cases, genetics, not environment, determines caste, and Keller thinks that there may be many more such examples among social insects. But he says it's taken a decade for his colleagues to accept this idea. "Many people are bad scientists," he says bluntly. "When they get data that doesn't fit, they do everything they can to throw out the interesting stuff."

    Still, some others suspect that environmental factors, not genetics, are the key to caste most of the time. "It's up in the air whether [genetic caste determination] is extremely rare or just occasional," says evolutionary biologist Francis Ratnieks of the University of Sussex in the United Kingdom.

    Robotic stand-ins

    Some of his ideas defy current wisdom, but Keller, who is well established in an enviable position in Lausanne, is not worried about controversy. His office looks out on deep blue Lake Geneva, with the Alps as a backdrop. He's never had a Swiss National Science Foundation grant proposal rejected, and with the money he has in hand right now—including €2.5 million for 5 years from the European Research Council—he can afford to take on risky projects and keep his 15 postdocs and students employed indefinitely. Just 50 kilometers away is his chalet, where he spends winter weekends skiing, sometimes with his two teenage children or members of his lab. Colleagues rave about how fun he is at meetings; several have photos of him dancing on a bar table late one night in Japan.

    That solid standing has helped Keller start unusual collaborations. Although his lab is adept at manipulating ants to understand their society, the researchers have been hard-pressed to come up with ant experiments that address how sociality arises in the first place. So a decade ago, Keller went across campus to the Swiss Federal Institute of Technology in Lausanne and teamed up with roboticist Dario Floreano. Together, they and their students harnessed robots and computer simulations to look at the early evolution of cooperation and communication. "The robots are really valuable for understanding limits and constraints on evolutionary processes," says Melanie Moses, a computer scientist at the University of New Mexico, Albuquerque.

    Each robot is built with a slightly different neural network connecting its infrared and vision sensors to its moving wheels. These variations, surrogates for genetic differences, cause the robots to have slightly different behaviors. The researchers score the robots for how well they do retrieving large and small disks, and the most efficient ones are selected to continue and "reproduce": In each successive round, the researchers increase the proportion of the best robots in the mix, adding clones that have slight variations reflective of what might happen during reproduction.

    At first, movements and diskretrieval are random, but within 150 generations the selection process leads to robots that excel at working together to retrieve large disks, the team reported in 2010. The next year, they demonstrated that the robots even evolve altruism toward their closest relatives. The work showed how easy it is for cooperation to emerge in organisms lacking much of a brain.

    It doesn't take much of a brain to evolve communication either, according to other work by the team. In these experiments, the robots have the ability to flash blue or red when they encounter "food." Within 10 generations, different sets of robots adopt different signaling strategies—flashing one color to signal that they have found food, or using combinations of colors to signal food or no food. The simple system is more efficient when there is no competition. But when competing robots try to read the signals, the complex codes are harder to break and thus more useful, the Lausanne team reported on 3 January 2012 in the Proceedings of the National Academy of Sciences. "[Keller] showed how a complex strategy can evolve and could beat simpler strategies," Moses says.

    Floreano and Keller are coming up with ever more complicated ideas to test with robots and simulations, but Keller is most excited now about the new ability to follow individual ants in colonies. Even with genomics, robots, and a new tracking system under his belt, however, Keller says he still feels like a rolling stone whose direction is not yet determined. "I hope in 5 years I'm starting to do new stuff that I am not even thinking about today."

  6. The Private Lives of Ants

    1. Elizabeth Pennisi

    Six years ago as a graduate student of Laurent Keller, Danielle Mersch developed a 24/7 surveillance system that promises to revolutionize their team's study of collective behavior in fire ants. Online this week in Science, they describe using this system to reveal that ants cluster into cliques.

    Ants at a glance.

    Ants wearing barcodes are tracked by overhead cameras that record the ants' positions in their homes (bottom).


    Danielle Mersch's ants have no privacy. Six years ago as a graduate student in Laurent Keller's lab at the University of Lausanne in Switzerland, she had wanted to study circadian rhythms in ants but hadn't wanted to spend all day and night keeping track of them. So she and Keller teamed up with Alessandro Crespi, an engineer at the Swiss Federal Institute of Technology in Lausanne to develop a 24/7 surveillance system that promises to revolutionize their study of collective behavior in these insects. "It's groundbreaking technology," says Iain Couzin, a system biologist at Princeton University.

    Online this week in Science (, Mersch, Crespi, and Keller describe using this system to reveal that ants cluster into cliques: Nurses that tend the queen and brood hang out in one part of the colony, while foragers stick close to the nest entrance, with the two groups rarely interacting. Some organization was expected, but "I was surprised that they have such a strong spatial structure," Mersch says.

    The paper's most significant advance, however, may be the setup itself. Mersch struggled to find a suitable marker that would let a computer follow each ant and take data frequently. Many ant researchers paint colors on the ants' abdomens to uniquely identify each one, but colors are hard to incorporate into an automated tracking system, especially one that has to operate in the dark. Radio tags don't provide enough spatial resolution, Mersch says. Other automated trackers record data at long intervals, leaving gaps in coverage.

    Mersch settled on a set of 225 barcodes developed for video game developers and glued a barcode to each ant's back; the ants seem to get used to their signboards. High-resolution cameras mounted above the ants' home—a shoebox-sized nest box and foraging area connected by a tunnel—is synced with infrared lights that flash on twice a second. The digital images are immediately relayed to a computer, where software analyzes each ant's position relative to other ants and computes their trajectories. Thus, Mersch can watch ants move in real time or days later.

    Typically, she starts with 200 ants and has the system observe them for 40 days. So far, she and her colleagues have spied on five species of ants. "It's very important information when you want to investigate the dynamics within societies," says systems biologist Guy Theraulaz of CNRS, the French national research agency, in Toulouse.

    About one-third of the colony works with the brood in one corner of the container as nurses; another third forages. Foragers turn out to have a daily rhythm, temporarily leaving the nest by night, but nurses don't. Many of the rest may be housecleaners. Over time, nurses are likely to become housecleaners and, eventually, foragers, moving further out into the world as they age. "There's a clear behavioral trajectory," Mersch says. Next, she plans on manipulating the social structure of the colony, say, by removing all the nurses, to see which ants take their places.

    "Until now, biologists have looked at the colony level," Couzin says. "Now they can look at the life history of every individual in the colony. This offers a remarkable way of relating individual properties to collective properties."

  7. Grand Challenge: Undergraduate Teaching

    Transformation Is Possible if a University Really Cares

    1. Jeffrey Mervis

    The same attention to scientific detail that led to his Nobel Prize is helping Carl Wieman improve how undergraduates learn science.

    The way that most research universities across North America teach science to undergraduates is worse than ineffective, says Carl Wieman. It's unscientific.

    A Nobel Prize–winning physicist turned science educator, Wieman doesn't understand why institutions of higher education would disregard decades of research showing the superiority of student-centered, active learning over the traditional 50-minute lecture. Using that outdated approach, he says, means universities are squandering talent at a time when U.S. higher education is being criticized for not turning out enough science-savvy graduates to keep the country competitive.

    Engaged instruction.

    Carl Wieman uses active learning tools to teach an undergraduate course at the University of Colorado in 2001.


    Wieman has spent the past 15 years applying the science of learning to how undergraduate science courses are taught. First at the University of Colorado, Boulder, ( and, more recently, at the University of British Columbia (UBC), Vancouver, in Canada (, Wieman and his colleagues have made impressive strides in changing how individual faculty members teach. Those changes, within individual courses, have translated into big improvements in student learning.

    Those courses are offered by academic departments, which are his real target. Departments define the reward structure for faculty members through their authority to hire, promote, and grant tenure, he says. So the best way to sustain improvements in teaching and learning is to get departments to buy into the need to change the courses that they offer. And that's begun to happen at UBC, one of Canada's elite universities.

    Wieman's passion for the subject, combined with his stature as a Nobelist, has focused national attention on the high attrition rate among students who declare an interest in earning a degree in a STEM (science, technology, engineering, and mathematics) field. It's one of the biggest impediments to any effort to train more scientists and engineers. "I think Carl, more than anybody else, put a spotlight on the need to improve undergraduate education," says Subra Suresh, who last month stepped down as director of the National Science Foundation (NSF). "It wasn't a surprise to universities, but his work has highlighted the problem."

    Colleagues also laud Wieman's rigorous approach to reform. "I have an incredible amount of respect for his deep commitment to the evidence," says Susan Singer, head of undergraduate education at NSF and a national leader in reforming undergraduate biology education. "Carl is someone who digs in and really wants to know."

    Notwithstanding his success at Colorado and UBC, Wieman has made much less progress toward another of his goals: overturning an academic culture that values research over teaching. Working in the White House Office of Science and Technology Policy (OSTP) as associate director for science, Wieman was the de facto science education czar for the Obama administration. But his 20 months on the job taught him just how hard it is to change prevailing attitudes within U.S. higher education.

    While at OSTP, Wieman floated the idea of requiring universities to collect and disseminate information on their teaching practices to remain eligible for federal research dollars. The policy would be a stick to get universities to pay more attention to teaching, he reasoned.

    "There's an entire industry devoted to measuring how important my research is, with impact factors of papers and so on," Wieman says. "Yet, we don't even collect data on how I am teaching. It receives no attention. … If everything about teaching remains hidden, then universities can avoid having to devote anything more than minimal effort to doing it well. They can instead spend most of their time and money on research."

    Wieman pushed the idea at numerous meetings with other government science officials and academic leaders. But they recoiled in horror at the prospect of what they viewed as another unfunded federal mandate. They prefer a 5-year effort begun last year by the 62-member Association of American Universities that aims to create a voluntary "framework" for improving teaching practices that institutions can adapt to their own situation and implement at their own pace.

    "I'm very supportive of improving undergraduate STEM teaching," says Francis Collins, director of the National Institutes of Health (NIH), which spends more money on academic research than any other federal agency. "But this struck people as the wrong pathway by which to achieve the desired outcome, and not very fair."

    As if taking on the nation's research establishment wasn't enough of a challenge, last June, Wieman received a sudden diagnosis of multiple myeloma. To deal with this health crisis, he abruptly resigned from his White House post and enrolled in a clinical trial at NIH using two experimental drugs. That treatment ended in January, and the 62-year-old Wieman says he's "happy and healthy."

    Wieman, who is leaving UBC but declined to say where he's going, has returned to the lecture circuit with an updated version of his standard talk, entitled "Taking a Scientific Approach to Science and Engineering Education." He's definitely not cowed by the prospect of taking a long, hard road toward his goal. In fact, his personal metric for any reform worth attempting is its ability "to generate significant opposition." Speaking at a session of the February annual meeting of AAAS (which publishes Science) in Boston, Wieman said that transforming undergraduate teaching "passes that litmus test."

    Giving reform a chance

    Wieman's personality and upbringing seem well-suited to a grand challenge like remaking undergraduate science education. Before deciding on a scientific career, he embraced a succession of passions, including chess and tennis, which for a time were all-consuming. "Monomaniacal pretty much describes me," Wieman confessed during a 2007 interview with the Nobel committee. "My view of everything is that you become good at something by focusing and working hard at it." Eventually, he recalls, "science [became] such an activity."

    That doggedness served him well in pursuing his Nobel Prize–winning research. In 1925, Albert Einstein, building on the work of Indian physicist Satyendra Nath Bose, deduced that cooling a gas of certain atoms should make all the atoms suddenly flop into the same lowest energy quantum wave. Such a macroscopic matter wave is known as a Bose-Einstein condensate. Some 70 years later, Wieman and Eric Cornell of JILA, a lab run jointly by the National Institute of Standards and Technology and the University of Colorado, Boulder, achieved one by employing magnets and lasers to cool rubidium-87 atoms to within a millionth of a degree of absolute zero. In 2001, the two physicists shared the Nobel Prize in physics with Wolfgang Ketterle of the Massachusetts Institute of Technology in Cambridge, who achieved a similar result with sodium-23 atoms.

    Wieman and Cornell at least had the advantage of knowing what a Bose-Einstein condensate would look like before they created one. In contrast, many faculty members might not recognize high-quality, student-centered learning because they may never have experienced it. Wieman admits that many notable scientists have thrived on a diet of traditional teaching practices and that the current rewards system at most universities gives faculty members little reason to try something different.

    "I'm certainly not one to dismiss the importance of research," Wieman says. "But people need to recognize how totally dominant the reward system is. There are a lot of faculty who feel, completely appropriately, that 'I could spend more time improving my teaching, but that's not what I'm supposed to be doing.' So you have to figure out a way for them to be able to improve their teaching without making a big sacrifice in their research activities."

    Wieman embarked on his quest to improve undergraduate education after pondering his own career as a professor and educator. And like the scientist that he is, he began by asking himself some basic questions. Why, he wondered, did students in his introductory courses do so poorly, and even regress, after he delivered lectures covering what they needed to know? Why couldn't he identify at the outset which graduate students were most likely to succeed? And why did most of them become productive scientists after a few years in his lab?

    Digging into the literature on teaching and learning yielded some insights. His graduate students had learned to think like scientists, he realized, by doing real science under the supervision of a world-class scientist. Developing expertise, he came to understand, is a slow and arduous process marked by repeated failures.

    "The apprentice model works pretty well in graduate school because the faculty member can see if the student is learning how to build a laser system, or write a paper, or give a professional talk," says University of Colorado, Boulder, physicist and education researcher Noah Finkelstein, who has worked closely with Wieman and now directs the university's newly formed Center for STEM Learning. "Those are things we actually want them to do. We give them feedback along the way, and we take in feedback from them and adjust our mentoring. But that system is just too costly at the undergraduate level."

    Instead, faculty members must interact with hundreds of students in a large hall. Most choose to do that via a lecture. But research has shown that most students cling to their misconceptions even after sitting through a brilliant lecture.

    What works better than lectures and homework problems, according to numerous studies, is having students work in small teams with instructors who can help them apply those basic concepts to real-life situations. But what's the best way to implement active, student-centered learning? The answer, Wieman decided, lay in melding it with the concept of deliberate practice.

    That idea, developed by psychologist K. Anders Ericsson of Florida State University in Tallahassee, treats the brain as a muscle that must be exercised to perform at its peak. It's how a novice becomes an expert, whether in music, sports, or science. "We have learned that complex expertise is a matter not of filling up an existing brain with knowledge, but of brain development," Wieman says.

    Deliberate practice, Wieman wrote in the fall 2012 issue of Issues in Science and Technology, "involves the learner solving a set of tasks or problems that are challenging but doable and that involve explicitly practicing the appropriate expert thinking and performance." The teacher, or coach, offers appropriate incentives to encourage students to master the necessary skills, as well as continuous feedback to help them remain on task. As with any sport, he notes, "[t]housands of hours of deliberate practice are typically required to reach an elite level of performance."

    The two concepts created an intellectual framework around which to transform undergraduate science. "Just as we have physics principles, here are the principles that work, and they are consistent with what others had done," Wieman says. "It also allows you to go into disciplines where there hadn't been much work done, like oceanography, and make some generalizations. It's very much like science itself."

    In a 2011 paper in Science, Wieman and his colleagues describe the power of active learning and deliberate practice. The instructor for one section of an introductory physics class for engineers at UBC used these principles, while the other instructor delivered the normal lectures. The first group of students scored more than twice as high on a multiple-choice test of the material covered than did those in the control group.

    "The results were so dramatic from this relatively modest experiment that the entire [physics] department had an epiphany," remarks Simon Peacock, UBC's dean of sciences. "It sent them a clear message: Wow, we can actually teach better."

    Wieman says that active learning and deliberate practice is now the norm in 99 UBC courses enrolling 31,200 students. Many are introductory courses taken by freshmen and sophomores who are still uncertain of their major field of study. "We have substantially changed more than half of the math and science courses a UBC student in the college of science will take in their first 2 years," Wieman says, citing results from a recent survey of how faculty members have changed their teaching practices since the Carl Wieman Science Education Initiative was launched in 2007.

    "We've hit it out of the park with earth and ocean sciences," one of seven departments that are part of the university-funded initiative, Peacock says. "I will declare them to be a success."

    Wieman believes that deliberate practice can also help students in primary and secondary school who, for whatever reason, are ill-prepared for success in STEM subjects. His efforts have helped resolve "a huge controversy," says NSF's Singer, over whether the vast majority of students are capable of doing high-level math and science.

    "Having Carl stand up and say we should stop doing STEM talent selection and start doing STEM talent development completely changes the nature of the conversation," says Singer, on leave from her post as a biology professor at Carleton College in Northfield, Minnesota. "It's really a question of how you structure the learning environment. And his work has shown that active learning strategies are more effective."

    From my way to the right way

    What does it take to transform an undergraduate science course? Wieman's approach relies heavily on a cadre of science teaching and learning fellows, who are typically postdocs. At its height, the Colorado initiative employed a dozen such fellows; at UBC, the number peaked at nearly two dozen.

    The fellows are trained in the many steps needed to transform a university lecture course—steps that faculty members are unlikely to take on their own, either out of ignorance or because they simply don't have the time to do what's needed. Katherine Perkins, who directs both the science education initiative at Colorado and the related PhET project (, which has created thousands of research-based simulations of physical phenomena, calls the teaching and learning fellows "engines of change."

    Meeting for the first time with a faculty member, a fellow might start by asking what the faculty member wants students to know how to do at the end of the course. That's a more useful metric than asking what a student "should understand," explains Beth Simon, director of the Center for Teaching Development at the University of California, San Diego, who spent the 2007 to 2008 academic year at UBC as a fellow in the computer science department before returning to UCSD.

    Once the faculty member articulates the real goals of the course, those skills are converted into learning objectives. The next step is to write up multiple-choice questions aimed at helping students achieve each learning objective. The so-called clicker questions (the name comes from the electronic device that students use to record their answers) usually focus on common student misconceptions about the concepts.

    The questions become the basic curriculum for the course. But getting from skills to clicker questions can be difficult. Simon figured that the final exam would provide a useful guide to what students were expected to learn. Instead, instructors would admit that they didn't really know what concepts some test questions were meant to measure, she says, and that other questions covered concepts not central to the course.

    Most courses come with only a three- or four-sentence description in the syllabus. That brevity gives whoever is teaching the course a lot of leeway. Some faculty members have been teaching the same course for years, Simon says, and for them, "learning outcomes were a nonstarter. 'I teach 101 my way,' they would say." In contrast, some courses are "owned" by the department and a consensus exists on what students are expected to know regardless of who is teaching the course.

    A transformed course typically begins not with a lecture but with a clicker question. Students gather in small groups to discuss it, and a fellow assigned to the course circulates through the classroom to guide the inquiry process. Once the students have punched in their answers, the faculty member might offer a microlecture aimed at correcting their mistakes and filling in gaps in their knowledge. Once the concept is clear, the class moves on to the next clicker question.

    Students taking transformed courses are usually more active than in a typical lecture class. Faculty members need to remind students regularly why they will not be lecturing, Simon says, as well as explain the importance of peer instruction. To get the most from the class time, students are assigned outside reading and turn in homework that measures their understanding of the material.

    Some students are uncomfortable with this approach—even if it's more effective. "I remember getting an evaluation from one [UCSD] student who had just finished my course," says Simon, a pioneer in the use of peer instruction within her field. "I loved it. It read, 'I just wish she'd have lectured. Instead, I had to learn the material myself.' "

    The increased student engagement in a transformed course is music to the ears of the average faculty member. "Most faculty want their students to learn more," says Perkins, whom Wieman hired in 2003 as one of the initiative's first teaching fellows. "They look at the final exam, sigh, and say, 'Why did only 60% get that question right?' " Simon adds, "If they can have more fun, they will choose to use these methods."

    A department should plan on spending about 5% of its budget for 5 years to transform its courses, Wieman says. Lesser amounts are required to sustain progress, he adds, although new faculty members must be trained and existing faculty members need ongoing support and, occasionally, a sympathetic ear. At Colorado, for example, departments competed for grants of roughly $600,000 to $800,000 each. UBC's $10 million commitment to the initiative allowed Wieman to double the size of departmental awards, and a more recent $2 million donation from David Cheriton, a professor of computer science at Stanford University, is fueling reform within the math and computer science department.

    Wieman's campaign to transform departments isn't the only game in town. Finkelstein's new center at Colorado, funded by an NSF planning grant, is supposed to serve as a focal point for some 75 STEM-related activities on campus. And Colorado's Perkins hopes that NSF will put up several million dollars for a Web site to help faculty members use the PhET simulations that she and others have created and to study their impact on teaching and learning.

    But money remains tight. Wieman says he can't afford to conduct the rigorous, outside assessments that normally accompany NSF-funded reforms because he feels that institutional funds should redound to the benefit of the institution. However, the dearth of peer-reviewed publications has led some scientists to question what Wieman's Colorado and UBC initiatives have accomplished.

    "When people ask what we've done," Finkelstein says, "and I say we've shifted institutional identity and culture, half the time their response is, 'Wow, that's terrific.' But the other half say, 'So all you've done is talk?' "

    Wieman himself offers a frank answer when asked whether he expects the UBC reforms to stick. "That's why you do research," he says. "This was a one-time intervention. And people have a right to wonder what will happen next.

    "I'm more optimistic than I was a year ago," he adds, "because people who we thought weren't interested are now saying, 'Look, I made this change and I'm thinking of doing more.' But I won't give you good odds that they will still be doing it in 10 years."

    Carrot or stick?

    In 2010, Wieman decided to come to Washington for the chance to influence undergraduate science education on a national scale. "My top priority at OSTP was to improve undergraduate education," he says. "We know what to do that will help students learn more and be more successful and how to get a broader group of students doing it."

    While there, Wieman came up with his simple, market-driven first step: Require universities to compile and release data on their teaching methods as a condition for receiving federal research funds. As students began using the data released by universities to help choose a college, he reasoned, universities would feel compelled to improve their teaching practices in order to attract the best applicants. "If an agency were to require every grantee to provide this information," he says, "then the next year teaching would look completely different because somebody is looking at it."

    Wieman promoted the idea tirelessly in meetings with his government colleagues as well as the presidents of several leading research universities, seeing it as a painless way to propel reform. But they pushed back hard. It's hard to define particular teaching practices, they told Wieman. Self-reported data are unreliable, they added, and collecting such data would be a burden. Last April, the presidents of several prominent universities even wrote a letter to then–White House Chief of Staff Jacob Lew in an attempt to head off Wieman's proposal.

    A few months later, Wieman was gone. But he hasn't changed his mind one iota, and he says that none of the community's objections are valid.

    For starters, he says, colleagues at UBC and Colorado have created a questionnaire that collects such data and requires only a few hours of effort by an entire department—"a tiny amount compared to what is spent in a single faculty meeting," he snickers. Universities have no incentive to game the system, he adds, because students would soon expose any institution that had submitted bogus information. And he scoffs at the idea that tracking a fundamental purpose of a university could be regarded as a "burden."

    Part of their objections, he speculates, is that the data could prove embarrassing. "Educational transparency is a threat to their status," he argues. "Maybe it won't make them look so good."

    NIH's Collins says that's not the reason he prefers a voluntary approach. "I know that Carl is skeptical universities will do it on their own," he says. "But I have yet to be convinced that they won't. I don't know that all universities will want to participate. But I think there will be some who would say, 'Yeah, we believe in this. It's the right thing to do.' "

    Government officials and university leaders typically defend the value of federally funded research by citing its role as an engine of economic growth. In the case of biomedical research, they also note its potential to save lives. But Wieman doesn't think those arguments really address the growing clamor from the public and politicians for universities to show that an undergraduate education is worth the rising cost of tuition. That skepticism, he says, has also fueled a decadelong assault by many state legislatures on their flagship public universities.

    A more effective response, Wieman says, would be for university presidents to emphasize how research can lead to better teaching. "I think the solution is to show that you can really use that research expertise to improve education," he says. "Deliberate practice and other approaches is calling on, and demanding of, the research expertise embodied by that faculty."

    "If you pitch that message," he continues, "then suddenly it becomes clear how having a great research university translates into better education for students in my state. Right now it's not worth the investment, because it's not happening. But it could."