# News this Week

Science  14 May 2010:
Vol. 328, Issue 5980, pp. 798
1. Fusion

# ITER Cost Estimates Leave Europe Struggling to Find Ways to Pay

1. Daniel Clery*

For the past 2 years, officials involved in the ITER fusion reactor project—poised to begin construction in France—have muttered darkly about huge increases in its estimated cost. Now, finally, we have some hard figures: Last week, the European Commission, the executive of the European Union, released a memo announcing that Europe's contribution to building ITER (45% of the total) would amount to €7.2 billion, 2.7 times the original estimate. Crucially, E.U. fusion funding up to the end of 2013 is short by €1.4 billion.

Senior fusion researchers in Europe who spoke with Science were astonished that the commission proposed no solution to the problem apart from asking E.U. member states for more money. With Europe's wider economy in turmoil, this funding crisis must be resolved before the mid-June ITER council meeting when the other partners expect to give the project's design, cost, and schedule final approval. (The June meeting is also expected to approve the appointment of a new director general for ITER, reported to be Osamu Motojima of the National Institute for Fusion Science in Toki City, Japan.)

ITER aims to show that fusing hydrogen nuclei inside a huge and complex reactor called a tokamak can produce enough excess energy to be a viable energy source. When the seven partners in the project—China, the E.U., India, Japan, South Korea, Russia, and the United States—signed the project into existence in 2006, they based their cost estimates on the design that had been adopted in 2001. Soon after the new ITER staff began work at the chosen site of Cadarache, it became clear that those estimates were far too low.

Under ITER's unusual funding arrangement, each partner must manufacture and deliver a share of the reactor components to the site. How much it spends making them is its own concern. Each partner apart from the E.U. is responsible for 9% of the reactor's “value”; the E.U., as host, supplies 45%. All of the partners got a nasty shock when they looked in detail at what they had agreed to make. The United States, for example, which had estimated its share would be $1 billion, is now expecting a bill for up to$2.2 billion.

For the E.U., with its proportionately larger share, the inflated cost is causing serious problems. In documents accompanying last week's memo, the commission blames the increases on a number of factors, including updates to the reactor design, the cost of setting up the organization to manage its procurement, the added complexity of having seven partners (there were only three in 2001), and extra resources for quality control. When the scale of the increases became apparent, the E.U. pushed for the construction schedule to be stretched out and called for inquiries into cost estimation and management structures (Science, 27 June 2008, p. 1707). Europe's stance has frustrated the other ITER partners. “Even if the budgetary situation is severe, it's important that every country properly fulfill its obligations under this international agreement,” says Yoshiyuki Chihara, director of the International Nuclear and Fusion Energy Affairs Division at Japan's Ministry of Education.

Last week's memo spelled out the budget tangle the commission has got itself into. The E.U. multiyear budget that supports ITER runs until the end of 2013. For the years 2012 and 2013, there is only about €700 million in the pot, but according to the memo the commission needs €2.1 billion during that time for procurements early in ITER construction. Commission officials looked into two possible ways of filling that gap. The first was a loan from the European Investment Bank, an E.U. institution that lends to development projects within Europe and, increasingly, large-scale R&D efforts. But the memo dismisses this option because it says there is no identifiable income stream to repay the loan—a view that one fusion researcher called “totally silly.” The second option is to transfer funds from other budget areas, but this too is rejected. “The budget is very tight; … there is now hardly any space for additional shifting of money between budget lines,” says Herbert Reul, a member of the European Parliament and chair of its industry, research, and energy committee.

In summary, the commission says “a sustainable solution will require a clear financial commitment for the life of the project and a commitment to finance any overruns outside that framework, by member states.” Researchers who spoke with Science say the commission got itself into this mess and now wants the E.U. member states to bail it out. “Europe is hugely obligated. It went out on a limb to lead the project. … Having taken it on, it has to find a way,” says one who asked not to be identified. Says another: “If the E.U. doesn't do everything to regain the confidence of the other partners, it is lost.”

• * With additional reporting by Dennis Normile.

2. AIDS Research

# Bioethicists Assail a Celebrated TB/HIV Treatment Trial

1. Jon Cohen

A heralded clinical study in South Africa that assessed the best treatment strategy for people infected with HIV who are receiving drugs for tuberculosis should never have been done, argue two bioethicists in an online posting published 5 May by the Hastings Center, a bioethics research institute based in Garrison, New York.

The study, Starting Antiretroviral Therapy at Three Points in Tuberculosis (SAPIT), began in June 2005 and was stopped in September 2008 after an interim analysis found that people who delayed starting anti-HIV drugs until they completed a 6-to-8-month course of anti-TB medication had twice the risk of death. The prominent research group that ran the 642-person study immediately reported the findings, which led the World Health Organization (WHO) to strengthen its guidelines; a full report appeared in the 25 February 2010 issue of The New England Journal of Medicine (NEJM).

Bioethicists Sean Philpott and Udo Schüklenk branded the study “deeply flawed,” arguing that it violated the Declaration of Helsinki as it “caused foreseeable harms and preventable deaths for a substantial number of impoverished and poorly educated South African trial participants.” They further lambasted the South African ethics committee that approved SAPIT, contending that its failure “to recognize the clinical, ethical, and legal deficiencies in this study is shameful.” They also noted that U.S. co-authors shared “the academic glory” but did not submit the study protocol to their institutions for ethical review. Philpott is based at Union Graduate College in Schenectady, New York, and Schüklenk, who works at Queen's University in Kingston, Canada, is co-editor of Bioethics.

The head of SAPIT, Salim Abdool Karim of the University of KwaZulu-Natal in Durban, says he is stunned by the criticisms. “When I looked at it, my head fell down,” says Karim, an epidemiologist who also directs the Centre for the AIDS Program of Research in South Africa, which ran the study. At the trial's start, says Karim, many clinicians worried about the dangers of simultaneously treating HIV and TB for three main reasons: TB antibiotics can weaken the effectiveness of antiretroviral drugs; ARVs can be toxic, leading people to stop taking all medicines; and mixing the two treatments increases the risk of a life-threatening disease called immune reconstitution syndrome. “The only reason we got involved in the study is because it was so obvious to us that we had no idea what we were doing with these patients,” says Karim, whose country has more HIV/TB coinfected people than any in the world. “When you put them on ARVs, they died. When you didn't, they died. We were at sea.”

Karim notes that the 2003 WHO guidelines, Scaling Up Antiretroviral Therapy in Resource-Limited Settings, explicitly acknowledges the confusion. “The optimal time to initiate ART in patients with TB is not known,” the guidelines state. In particular, complications from treating TB and HIV simultaneously are most acute during the intensive initial phase of TB drugs, which lasts 2 to 3 months. Thus the guidelines suggest that “deferring the start of ART may be reasonable in a variety of clinical scenarios.” They advise, however, that patients with advanced disease—defined as fewer than 200 CD4 white blood cells per milliliter of blood—“be started 2 weeks to 2 months after the start of TB therapy.”

The critics' main complaint concerns these severely ill people. In the SAPIT study, 213 people with an average CD4 count of 140 were randomized to delay ARV treatment until completing TB therapy, and 429 others with an average of 150 CD4s started ARVs either within 4 weeks of TB treatment or within 4 weeks of ending the initial intensive phase. The delayed arm had 27 deaths versus 25 in the much larger “integrated” treatment group. All but eight deaths occurred in people with fewer than 200 CD4s, and given that some people died who received ARVs early, the critics calculate that at least 10 “preventable” deaths occurred in the delayed arm. “It is true that the question of when to initiate ARV in TB patients was an open question,” says Schüklenk. “To move from there to the proposition that it was acceptable to delay ARVs in patients even with very low CD4 counts by up to 9 months is false.”

Karim points out that one of the most surprising findings of the study is that there was no significant mortality difference between the integrated and delayed arms until well after people stopped TB treatment. This, he says, underscores why so much confusion existed before the SAPIT results. “The deaths are really occurring past 10 months,” he says, noting that TB doctors typically have stopped seeing the patients by then and thus would never know whether early use of ARVs was critical.

Karim further notes that the SAPIT study design, in keeping with WHO guidelines, allowed “highly trained physicians” to start ARVs based on their judgment about an individual's needs. But Philpott and Schüklenk don't believe that occurred. “We really don't think those individuals randomized to the delayed arm were receiving the individualized care,” says Philpott.

Opinions of clinicians and public health officials who were not involved with the study vary widely. “I think most medical observers familiar with current HIV/AIDS research practice in Africa, with the rapid evolution of WHO guidelines, and with programmatic work on HIV/AIDS in low- and middle-income countries, would consider these criticisms unjustified,” says Kevin De Cock, the former director of WHO's HIV/AIDS program and now head of the Center for Global Health at the U.S. Centers for Disease Control and Prevention. “The Hastings piece is written very much from the perspective of what we know now, not how we viewed practice in 2003–4 when discussions about how to handle HIV-associated tuberculosis led to studies such as this.”

Some South African clinicians who treat coinfected HIV/TB patients agree with the critics. “I would never have allowed one of my patients or a family member to enter this study—it was just too dangerous,” says Francois Venter of the University of the Witwatersrand in Johannesburg. Venter stresses that coinfected patients with fewer than 200 CD4s can get sick and die in a day or two. Gary Maartens, a clinical pharmacologist at the University of Cape Town, adds that had the study limited enrollment to people with between 200 and 500 CD4 cells, it still would have answered an important question about timing of ARVs. “We did not need a clinical trial to tell us that deferring lifesaving therapy for 8 months in patients with advanced disease results in higher mortality,” says Maartens.

Still other HIV/TB specialists say the issue of including people with CD4 levels under 200 is more complex. “It's all too easy to say this shouldn't have been done,” says William Burman, who directs the HIV clinic at Denver Public Health in Colorado and specializes in designing HIV/TB clinical trials. Although Burman says he “had concerns” about the SAPIT study's design and that the results from people with the lower CD4 counts did not surprise him, he finds the Hastings critique “overly negative.” Despite WHO guidelines, many TB clinicians in poor, high-burden countries had little experience with ARVs when the study began and were reluctant to treat patients with them. “SAPIT has clarified that matter and will surely save many, many thousands of lives,” he says.

As for the criticism of the ethical review process, two U.S.-based co-authors of the NEJM paper, Gerald Friedland of Yale University and Waafa El-Sadr of Columbia University, say they contributed only to the writing of the report. Neither helped design the protocol, saw patients, or provided financial assistance. The U.S. Department of Health and Human Services Office for Human Research Protections does not require institutions to conduct reviews of protocols for co-authors under these conditions. Philpott counters that in the paper, Karim also notes that he has an affiliation with Columbia. Karim says his part-time, adjunct appointment mainly involves teaching and supervision of postgraduate students.

Friedland and El-Sadr, both highly respected HIV/AIDS researchers, strongly object to the SAPIT criticism. “I'm flabbergasted,” says El-Sadr. “This will be cited as one of the landmark studies in HIV/TB.” Friedland adds that the South African ethical review was done by people with proper credentials, and he says WHO's 2003 guidelines were based just on observational data, a much lower standard of proof than a randomized, controlled clinical trial. “Now the guidelines can be very clear and firm,” he says.

The debate likely will continue over the next few months. In July, Bioethics plans to run an extended version of the critique that appeared on the Hastings Center Web site. Many researchers also expect that NEJM will run letters to the editor and a reply in the next few weeks.

3. ScienceNOW.org

# From Science's Online Daily News Site

How Jupiter Got Its Stripes Jupiter and other gaseous planets are covered from pole to pole with stripes. But astronomers aren't exactly sure how they got there. Now a team of physicists reports that Jupiter's stripes may be produced in part by tides, a result of the gravitational tugging of its 60-odd moons. To back up that idea, they've performed an experiment to show that tides can produce just such stripes, at least in a very simple mockup of a gaseous planet.

Bonding With Offspring Grows New Neurons in the Mouse Brain For a father to truly bond with his children, he needs to grow some new gray matter. At least that seems to be the case in mice. A new study shows that when a mouse father nuzzles his pups, he develops new neurons that help him remember—and protect—those offspring later in life. The results suggest that in mice, and perhaps in humans, young babies and dads bond biologically in ways that can last a lifetime.

X-ray Vision, Without the Radiation X-ray–like imaging without the harmful radiation and cell phones with more bandwidth are closer to reality now that researchers have developed a novel type of lens that works with terahertz frequencies. The new lens is a metamaterial, an artificial material with a structure made from many tiny parts, and it could drastically expand what lenses can do.

Kon-Tiki, Bacteria Style New research reveals that, just like a rat clinging to a piece of floating wood after a flood, pathogenic microorganisms can set up shop aboard drifting bits of fish feces and other debris and ride them to far-flung destinations. Understanding more about the process should help public health officials develop better strategies to fight waterborne diseases and seafood contamination.

4. Astrophysics

# Herschel's First Images Spark Star-Formation Debate

1. Sarah Reed

When the Herschel telescope was launched into space last May, it promised new insights into our universe, detecting far-infrared and submillimeter wavelengths of radiation that have eluded previous missions (Science, 1 May 2009, p. 584). A year later, the European Space Agency instrument has passed its commissioning phase and is starting to deliver on that promise, scientists reported at the Herschel First Results Symposium in the Netherlands last week. “Some instruments are working even better than anticipated from the ground tests,” says astrophysicist Annie Zavagno of the Laboratoire d'Astrophysique de Marseille, France.

Herschel is the largest infrared telescope ever to be put into space, with a 3.5-meter mirror that's four times larger than any space-based predecessor. There are larger infrared telescopes in ground-based observatories, but Earth's atmosphere absorbs many infrared wavelengths that are vital for studying the births of stars. “The earliest stages of star formation are hidden from our view but can be revealed in the far-infrared and millimeter domain. This is why Herschel is so important for us and will open a new window on our understanding of star formation,” says Zavagno.

One of the results reported last week is already provoking debate. Astronomers had thought that galaxies have been forming stars at about the same rate for the past 3 billion years. However, Herschel observations of nearby galaxies suggest this isn't the case. The telescope detected strong emissions of infrared radiation in those galaxies, which indicates furious star formation, says Steve Eales of Cardiff University in the United Kingdom, a joint principal investigator on the Herschel ATLAS project. Eales and his colleagues conclude that in the past many galaxies were forming stars at 10 to 15 times the rate that we see in our galaxy today.

However, cosmologist Carlton Baugh of Durham University in the U.K. calls this interpretation “controversial.” Baugh and his colleagues, including Cedric Lacey, also at Durham University, have been using computer simulations of how galaxies form to make detailed predictions about what Herschel should find. In a paper published online in March in the Monthly Notices of the Royal Astronomical Society, they argued that determining star-formation rates from Herschel observations would be complicated because the telescope can only detect infrared radiation powered by massive stars. To get an overall star-formation rate, scientists must estimate the number of low-mass stars by extrapolating from a relationship that describes the distribution of mass within a newly formed group of stars.

This relationship is typically assumed to be universal for all stellar regions, but Baugh and his colleagues argue that the merger of two galaxies can trigger star formation in which a different mix of stellar masses is produced, with a higher-than-usual proportion of high-mass stars. “Star formation like this can produce more emissions at the long wavelengths probed by Herschel for a modest amount of star formation,” says Baugh.

Astronomers may be at odds over the explanation of Herschel's observations of strong infrared emissions from nearby galaxies, but they are in agreement on a second observation: They can't explain it at all. Herschel spied an “impossible” star in the early stages of formation. It has a mass eight to 10 times that of our sun; such a massive star will emit intense ultraviolet radiation that should blast away the cloud of gas that is feeding its growth. Yet the star spotted by Herschel, found within the star-forming cloud RCW 120 in our galaxy, is still surrounded by an envelope of gas and is therefore expected to get even bigger. Astronomers have seen much more massive “impossible” stars before—the star Eta Carinae is thought to be more than 100 times as massive as the sun—but this is the first time that one has been observed at such an early stage of its formation.

Aside from studying the birth of stars, astronomers also hope to use Herschel to analyze the chemical composition of comets and planetary bodies and to examine the molecular chemistry of the universe. The mission team is reporting some success with the latter, as Herschel's HIFI spectrometer has detected ionized water vapor, the first time that this ion has been detected outside the solar system.

A special issue of Astronomy & Astrophysics dedicated to Herschel's first results is due out this summer, but scientists predict that the instrument will fill numerous pages of journals for many more years to come.

5. Cancer

# Panel Finds Environmental Risks Neglected

1. Jocelyn Kaiser

A flap has erupted among health experts over a report last week from a presidential advisory group that concludes environmental pollutants are an underappreciated cause of cancer that is inflicting “grievous harm” on Americans. The report, from the President's Cancer Panel, has drawn fire both from an industry-leaning group and from the mainstream American Cancer Society (ACS) for distorting environmental cancer risks.

One of the report's authors, Margaret Kripke, an immunologist and professor emeritus at MD Anderson Cancer Center in Houston, Texas, acknowledges that the text is “deliberately thought-provoking” but defends its scientific rigor. “Part of the issue here is to stimulate a call to action, and you don't get that by having a bland” report, Kripke says. She is one of the two members of the authoring panel; the other is LaSalle Leffall, a surgeon at Howard University. A third member, cancer activist Lance Armstrong, who did not contribute to the report, left the panel in 2008 and was not replaced.

The 240-page report, called Reducing Environmental Cancer Risk: What We Can Do Now, states that “the true burden of environmentally induced cancer has been grossly underestimated” and that Americans “are bombarded continually with myriad combinations” of carcinogenic pollutants. It reviews exposures from a range of sources, including hot-button ones such as the plastics ingredient bisphenol-A, medical x-rays, and even cell phones. Among other recommendations, it urges people to use headsets with cell phones and recommends that Congress and the president adopt the “precautionary approach,” which would require manufacturers to demonstrate the safety of chemicals before distributing them on the market.

The American Council on Science and Health, a nonprofit organization with industry support, wrote that “the report practically plagiarizes the work of anti-chemical activist groups” including the Environmental Working Group (EWG), an environmental advocacy organization in Washington, D.C. EWG's Richard Wiles was one of 45 witnesses—including many government and academic scientists—who spoke before the presidential panel in 2008 and early 2009.

Among those who object to the report's conclusions is ACS's emeritus chief epidemiologist, Michael Thun. In a response last week, Thun wrote that ACS had issued a report on environmental risks last year that raised similar concerns about untested chemicals, the risks to children, and hazards of medical imaging. But Thun finds last week's report “unbalanced” because of its “dismissal” of cancer prevention focused on lifestyle factors such as smoking and obesity that make a much larger contribution to cancer incidence. Thun also disputed the report's claim that estimates of pollution-induced cancer are “woefully out of date,” asserting that this statement “reflects one side of a scientific debate.”

That debate centers on the low estimates in 1981 from University of Oxford epidemiologists Richard Doll and Richard Peto, who found that occupational exposures and pollution cause about 6% of cancers. Boston University epidemiologist Richard Clapp, who testified before the presidential panel, acknowledges that he and a few other academics have contested the Doll-Peto figures and have been at odds with ACS on the issue for years.

Peto, for his part, says the report's criticisms of his and Doll's methodology are familiar and “factually untrue.” He says, “I don't think there's any good reason to revise our estimates upward.”

The President's Cancer Panel, created by the 1971 Cancer Act, is independent, although staffed by a small office at the National Cancer Institute. A science writer compiles testimony and the panelists' own input and writes the report, says the panel's special assistant, Jennifer Burt. The report does not go through peer review, but staff fact-check the information, she said. She defends the report as “very measured.”

Kripke says she's “disappointed” by ACS's criticisms. She says they're ignoring previous reports from the same panel emphasizing the importance of lifestyle factors. As for the Doll and Peto estimates, she says the view that they need updating is “an opinion.”

Activists are hoping the report will spur efforts in Congress to make U.S. chemical safety regulations more consistent with the precautionary principle. But it's not clear whether the proposal will gain much momentum. Its immediate destiny is to join a shelf full of previous annual reports from this panel.

6. ScienceInsider

# From the Science Policy Blog

Methane-trapping ice, which has frustrated the first attempt to contain oil gushing off the shore of Louisiana, may have been a root cause of the blowout in the first place. So says University of California, Berkeley, professor Robert Bea, who has extensive access to company documents on the incident. There were signs that drillers did encounter hydrates: About a month before the blowout, a “kick” of gas pressure hit the well hard enough that the platform was shut down.

Biomedical science will lose a longtime champion with the retirement this year of Representative David Obey (D–WI), the chair of the House Appropriations Committee. Obey, 71, announced that he will not run for reelection this fall after serving 21 terms in Congress.

National Science Foundation Director Arden Bement offered a sober assessment of likely congressional action on the 2011 budget after members of the National Science Board suggested that prospects looked rosy. “You should not be surprised if we don't get the president's request,” Bement said about the $552 million (8%) increase for NSF that President Barack Obama proposed in February as part of his overall$3.6 trillion budget. Bement also said “I won't be surprised to see us operating under a continuing resolution” until well after the November congressional elections, with funding frozen at current levels.

The most ambitious U.S. effort to assess environmental change on a continental scale has won final approval from the oversight body of the National Science Foundation. More than a decade in the making, the $434 million National Ecological Observatory Network (NEON) will establish 20 permanent monitoring stations to collect climate, environmental, and biological data on an ongoing basis. NEON will also include 40 temporary terrestrial sites and 46 aquatic sites For the full postings and more, go to news.sciencemag.org/scienceinsider. 7. Scientific Facilities # Sweden Bets on New Lab to Spruce Up Its Bioscience Future 1. John Travis The tree bears the name of a neighbor—Norway spruce (Picea abies)—but this conifer is central to Sweden's fiscal health. It feeds the nation's vigorous timber industries and by some estimates is economically the country's most important species. So it wasn't surprising when the Sweden-based Knut and Alice Wallenberg Foundation last year announced it would provide about$10 million for the sequencing of the Norway spruce's genome. But what is unexpected—or at least it would have been a year ago—is that this sequencing will largely happen in Sweden rather than being farmed out to other countries. Next week, officials in Stockholm will inaugurate a new building with cutting-edge DNA sequencing machines that by 2013 should produce a rough draft of the genome of the “Christmas” tree.

For Sweden, this building represents far more than a single research project. It's the most visible evidence to date of the Science for Life Laboratory (SciLifeLab), an unusual initiative spanning multiple Swedish research organizations and two sites, one in Stockholm and the other in nearby Uppsala. Rather than spread its money to as many scientists as possible, as it has traditionally done, the Swedish government will spend more than $75 million setting up SciLifeLab with the hope that it will become a technology-driven national life sciences center, comparable to organizations such as the Broad Institute in Cambridge, Massachusetts. “We want to make sure [the latest] technologies are available to all of Sweden. It's a deliberate decision to do things differently,” says geneticist Kerstin Lindblad-Toh, who is the director of SciLifeLab's Uppsala effort and has a position with the Broad Institute. The dream of creating a national research center in biology has been discussed for years by officials at the Karolinska Institute, Stockholm University, and the Royal Institute of Technology (KTH), all in Stockholm. But it only became a reality when the Swedish government recently offered major strategic research grants for several life sciences fields. The three institutions banded together last year to win the lion's share of the “molecular biosciences” money and then joined with Uppsala University, another grant winner, to form SciLifeLab. By 2011, the building in Stockholm should house more than 200 researchers. Aside from sequencing the Norway spruce, SciLifeLab intends to help Sweden take advantage of its formidable population databases and biobanks, such as the recently launched LifeGene project, which aims to gather tissue samples and track the medical histories of more than 500,000 Swedes. SciLifeLab “is what we really need to be competitive with other places and to analyze our own data,” Lindblad-Toh says. In a leap, Sweden will create one of Europe's biggest genome centers. There are much larger centers around the world—SciLifeLab's DNA sequencing output will be only about 40% of that of the Wellcome Trust Sanger Institute in Hinxton, U.K., for example. But Sweden is counting on an unprecedented marriage of genomics with proteomics to set it apart. Uppsala University and the Royal Institute of Technology are already home to a project, the Human Protein Atlas, that by 2014 intends to create antibodies to nearly every human protein. SciLifeLab “will be the only lab with antibodies to all human proteins and the ability to sequence human genomes. It's a unique niche,” says KTH's Joakim Lundeberg. Scientists say they will use antibodies from the Human Protein Atlas to examine subcellular locations of human proteins, for example. And there are plans to analyze the Baltic Sea ecosystem, notes KTH's Mathias Uhlén, who heads the Human Protein Atlas and Stockholm's SciLifeLab. Lindblad-Toh says the Uppsala side will add its strengths in comparative genetics and evolutionary biology to the mix. For now, the Norway spruce is center stage. No conifer has had its genome sequenced yet, in part because they typically are full of repetitive DNA sequences; Norway spruce and the human genome likely have comparable gene numbers, but the former is seven to 10 times bigger. “It's one of the largest, most complicated genomes to be sequenced,” says Pär Ingvarsson of Umeå University, who leads the spruce project. “We want to identify genes that control wood properties.” SciLifeLab's partnership with Umeå on the spruce genome is a first step toward convincing Swedish scientists outside Uppsala and Stockholm that the project is for them as well. “Most universities in Sweden are waiting to see if we can deliver useful infrastructure for them. It's up to us to prove we can,” acknowledges Uhlén. SciLifeLab may also help the scientific communities in Uppsala and Stockholm, traditional rivals, come together. The two SciLifeLab sites intend to submit many joint grant applications, and a single person may ultimately direct the overall effort. “For the first year, we will work as two separate organizations. We're working on getting closer,” says Fredrik Sterky, site director for SciLifeLab in Stockholm. “It's a lot more powerful if we can grow as one institute with two campuses,” Lindblad-Toh says. SciLifeLab represents a change from the past “dark decade,” when Scandinavian countries failed to use their wealth to fund strategic research efforts, notes Finland native Aarno Palotie, head of medical sequencing at the Wellcome Trust Sanger Institute. “Sweden woke up sooner than Finland or Norway.” Still, he concludes, SciLifeLab remains a gamble. The key question, asks Palotie: “Is the sum bigger than the components?” 8. The Laser at 50 # Taking Laser Science To the Extreme 1. Daniel Clery Europe wants to leap to the next generation of laser facilities with a 200-petawatt laser that will create new areas of research, and could rip open the vacuum. The first half-century of the laser's history has seen a constant push for higher power. Today, the Vulcan laser at the Rutherford Appleton Laboratory (RAL) near Didcot, U.K., fires pulses that have 10,000 times the power of all of Britain's electricity-generating stations added together. One of the world's most powerful lasers, Vulcan doesn't black out the entire country because its pulses are very short, less than a picosecond (10−12 seconds) in duration, so the energy of each pulse is a moderate 0.5 kilojoules. Vulcan is a large machine, but over the next few years a group of European countries wants to take lasers into the realm of international big science with a facility built around a device that can produce 200-petawatt (2 × 1017 watts) pulses, 200 times the power of today's best lasers. The Extreme Light Infrastructure (ELI) isn't yet a done deal, but there is considerable political and scientific momentum behind it. That's in part because the three countries leading the project are the Czech Republic, Hungary, and Romania—all recently joined members of the European Union. “There is political pressure from the new states and the E.U. to build [research facilities] in the new states,” says Wolfgang Sandner, director of the Max Born Institute in Berlin. If ELI goes ahead, it will be the most prominent science project in Eastern Europe since the fall of the Berlin Wall. Initially split into outposts in the three countries, leading up to one mammoth laser to be built by 2017, ELI will ultimately be the Swiss army knife of laser centers. Its superfast, high-power pulses will probe the atomic nucleus and watch electrons inside atoms and molecules. By colliding pulses with various targets, researchers plan to create other sorts of radiation—electrons, protons, ions, x-rays, and gamma rays—for use in everything from cancer therapy to nuclear physics. The ultrahigh power and intensity of pulses of ELI's final laser will produce electric fields so strong that they may alter and sense the texture of the vacuum itself, opening up new research areas for astrophysicists and particle physicists. According to quantum electrodynamics (QED), the vacuum teems with pairs of electrons and positrons that pop fleetingly into existence, briefly separate, and then recombine and disappear. The electric fields of ELI's pulses may be strong enough to pull these pairs apart before they can recombine, or at least feel the texture of this sea of virtual particles and test QED in a very direct way. We want to “drill a hole in the vacuum,” says ELI project coordinator Gérard Mourou, director of the Laboratory of Applied Optics (LAO) at Palaiseau, France. ## The power of three About 5 years ago, the E.U. called for ideas for international infrastructures to boost European research. Dozens of labs around the world already boast terawatt (1012 watts) lasers, and a handful, including Vulcan, can now reach petawatts (1015 watts). Mourou and others wanted to go even bigger. They put together a plan for a laser that would push current technology to its limits, into the hundreds of petawatts. Laser science “is ready to go to the next step, to a truly international laser infrastructure beyond the capability of a single nation,” says Sandner. Several European countries expressed interest in hosting ELI, but securing funding proved difficult until the three eastern countries realized they could apply for E.U. structural funds. These are grants given to less developed E.U. member states to build infrastructures such as roads, bridges, and hospitals, but they can equally well be spent on research facilities. Last year, the Czech Republic, Hungary, and Romania came up with a novel plan: They would become equal partners in the project and split it in three so that each would have a facility geared to a different branch of laser science. (Institutions in another 10 E.U. nations are also involved.) The three centers would have lasers with a range of powers between 1 and a few tens of petawatts; and the decision on where to put the final 200-petawatt laser would be put off for 2 years to give researchers more time to choose the best technology. Although laser scientists acknowledge that this makes the project more complicated, there are benefits, too. “There will be a slight increase in cost but a huge boost to local scientists [in each country],” says Sandner. If the structural funds are approved by early next year as expected, the three countries could begin pouring concrete in 2011, with a total price tag of about €750 million. “This is very, very significant. It's the first time a European infrastructure project has been built on the east side of the [former] Iron Curtain,” says physicist Marius Enachescu, who is deputy secretary of state in Romania's research ministry. The ELI facility in Hungary will focus on science using ultrashort laser pulses, just attoseconds (10−18 seconds) in length. Researchers began making attosecond pulses about a decade ago when they found that if they fired a femtosecond (10−15 seconds) laser pulse into a gas such as neon, they created higher order harmonics of the original frequency. By superposing these harmonics, they could create attosecond-scale pulses, which is just the time scale needed to discern the movement of electrons in an atom. Researchers hope Hungary's ELI facility will enable them to carry out “pump-probe” type experiments, in which one pulse sets an atomic process in motion, then a second snaps the action a moment later like a hyperfast camera. They say they will be able for the first time to image the position in time and space of both nuclei and electrons at the subatomic scale. The Czech branch of ELI will be a laser-based beamline facility. Many areas of science rely on beams of particles and high-energy photons from accelerators, synchrotrons, x-ray tubes, and radioactive sources. In 2000, researchers discovered that they could also generate many of these beams by firing high-intensity laser pulses into gas jets, thin foils, and other targets. This laser strategy can produce x-rays and gamma rays, as well as pulses of electrons, protons, and ions, with a brightness and pulse length that open up new experimental possibilities. “When a laser interacts with a target, all sorts of impressive things are created. The products often cannot be produced in any other way,” says John Collier, RAL's head of high-power lasers. One possible application this ELI branch plans to explore is cancer therapy with proton or ion beams. Such beams are extremely effective for treating deep-seated tumors, but to perform such therapy a hospital now needs a particle accelerator costing tens of millions of dollars—something few can afford. Laser physicists think they can accelerate particles in a much cheaper and more compact way. When they fire a high-intensity laser pulse into a plasma, the photons' magnetic field kicks electrons in the plasma forward and these then strike a foil target. As they emerge from the other side, they drag positive ions in the pulse's wake. Such acceleration works “much faster over a shorter distance” than traditional accelerators do, says Sandner. “It's in its very early infancy, but it points a way to the next step.” The third planned ELI facility, in Romania, aims to open up a new area of laser science by probing the atomic nucleus with beams that are ultraintense—focused so that they have the maximum power per unit area. “The laser power that exists now cannot be compared with the strength of the nuclear field,” says Enachescu. But he hopes that his nation's ELI outpost can change that. It's not clear yet whether ELI's laser beams alone will be able to excite a nucleus into higher energy levels, but physicists are developing other tricks that utilize those beams to accomplish the feat. One such scheme involves colliding a laser pulse head-on with an electron beam to produce an intense burst of gamma rays. “Then we will use that to disturb nuclei,” says Mourou. ## Probing the vacuum The fourth, and at the moment least-defined, part of ELI is the final 200-petawatt laser. The uncertainty is because planners are still weighing two rival methods to stretch current technology to this new power level. All high-power research lasers rely on amplifiers: pieces of an active lasing medium, such as glass doped with neodymium, that resemble a laser without the end mirrors. Just before a pulse is fired, the amplifier is pumped with light from another source to create a large number of excited atoms. When the pulse comes through, those atoms emit light in step with the pulse, amplifying it with extra photons. But at about a gigawatt, each pulse has so much power that it begins to damage the glass. Researchers got around this problem in the mid-1980s after Mourou and colleague Donna Strickland developed a technique called chirped pulse amplification (CPA), which reduces the peak power of a short, high-power pulse by stretching the pulse out in time, before amplifying it and compressing it again. ELI may need additional strategies to increase laser power. “We're getting to the limit of CPA,” says ELI Deputy Coordinator Georg Korn of the Max Planck Institute of Quantum Optics in Garching, Germany. The ELI team may consider an adaptation of CPA involving a special nonlinear crystal to transfer power from one stretched beam to another. This will be tested in a planned upgrade of Vulcan to 10 petawatts. Meanwhile, Mourou's LAO and other French labs are testing a different amplifier material, titanium-doped sapphire, by building a 10-petawatt laser. The ELI team must eventually decide which approach to back and whether to push for even higher power or simply build 20 10-petawatt lasers and combine the beams to make one of 200 petawatts. “We have to wait for this new technology to develop. Two-hundred petawatts is so advanced that there is a need for a demonstrator,” says Collier. Even if researchers achieve that power, they will still need high-intensity beams before they can explore the vacuum. Theorists calculate that an intensity of 2029 watts/square centimeter (W/cm2) will be needed to rend apart electron-positron pairs. ELI will likely be able to reach intensities of only about 1024 W/cm2, but “there are some clever ideas around, some of them not yet published,” says Korn. These include using laser pulses to create intense gamma rays and probing the vacuum with them. With enough funding, says laser scientist Donald Umstadter of the University of Nebraska, Lincoln, ELI should overcome any technical difficulties. “They've set ambitious goals to reach ideal conditions. If they do, it will be very exciting. Whenever you are going to the limits, you can expect interesting physics to emerge.” 9. Power # Laser Fusion Energy Poised to Ignite 1. Daniel Clery Once lasers have shown that they can spark fusion, what next? Projects in Japan, Europe, and the United States are working to build lasers with the necessary power and repetition rate to make energy from laser fusion a reality. In 1960, John Nuckolls, a physicist at Lawrence Livermore National Laboratory (LLNL) in California, was studying ways to compress and heat a capsule of hydrogen isotopes so that they would undergo fusion and generate energy. He looked at then-exotic technologies such as pulsed-power machines, particle accelerators, plasma guns, and hypervelocity pellet guns. Then in July he heard that Theodore Maiman had built the first successful optical laser at Hughes Research Laboratory in Malibu, California. Within days Nuckolls was working out how to implode his fuel capsule with a laser. Today, the direct result of Nuckolls's initiative is sitting in a cavernous hall 10 stories high and large enough to cover three football fields: the$3.5 billion National Ignition Facility. NIF was completed in 2009, and test firing of its giant laser has been going on since last summer. Shots aimed at igniting fusion fuel could begin later this year. Nuckolls—now a director emeritus of LLNL—and the rest of the fusion community are waiting anxiously to see if NIF's laser, which produces the world's highest energy pulses, really can ignite a tiny version of the sun. “NIF will prove that ignition takes place at laboratory temperatures, irrespective of the heating method,” says Hiroshi Azechi, director of the Institute of Laser Engineering at Osaka University in Japan. Laser, or inertial, fusion energy “needs ignition to be successful,” says Mike Dunne, director of the Central Laser Facility at the Rutherford Appleton Laboratory in Didcot, U.K. “There's a lot of belief. The question is when?”

Ignition alone, however, is not enough to prove that inertial fusion is a viable energy source. For all of its importance as a demonstration project, NIF would make a terrible power station. To give NIF the best chance of success, LLNL researchers built a laser designed to produce one-off pulses with the highest energy currently possible: 1.8 megajoules in a flash lasting about 20 nanoseconds. The tradeoff is that NIF can fire only two or three times per day. If everything goes according to plan and each fusion burn produces 100 megajoules of energy, three shots a day would produce about 3.5 kilowatts of heat—enough to power one typical American home. Researchers estimate that an inertial fusion power station would need to explode a fuel capsule about 10 times a second, a feat no high-energy lasers can now achieve.

Other obstacles to inertial fusion power include manufacturing the tiny fuel capsules filled with hydrogen at a rate of about a million a day; feeding them into the path of the laser beams 10 times a second; and building a vessel to contain the explosions and absorb the neutrons produced in fusion while converting their energy into heat to generate electricity. Even as they look forward to news from NIF, researchers at projects in Japan, Europe, and the United States are already exploring steps toward practical laser-driven fusion power, should the laser in Livermore succeed. Which projects offer the best hope, and whether all can survive tight funding, remains unclear.

## Pulse power

Key to any attempt at inertial fusion energy is building the laser. Industry has had decades of experience with high-power lasers of about 100 kilowatts, used for cutting, for example. Fusion could manage with 50 kilowatts on average, Dunne says, but it needs to be delivered not as a continuous beam but in very short pulses each with high energy. “The challenge is to marry short pulses with high average power,” says Dunne, a combination he thinks will take about 5 years to achieve.

High-energy lasers such as NIF's start with a relatively low-power beam and pass it through a series of amplifiers that steadily boost its power. An amplifier is a large piece of lasing material pumped with light from another source to create a population of excited atoms. When the beam passes through, the atoms emit photons in step with it, increasing its power. NIF's amplifiers are made from glass doped with neodymium; the pumping is done by thousands of xenon flash lamps powered by a bank of capacitors that can store 400 megajoules of energy. That's fine for NIF but not for a true laser-fusion power plant: Flash lamps must be cooled between shots and their capacitors recharged, and they are less than 1% efficient from wall plug to light. A power plant would need something at least 10% efficient—which, most agree, means solid-state laser diodes.

Laser diodes, developed for the telecommunications industry to send light down optical fibers, are “getting better all the time,” says NIF Director Ed Moses. A diode-pumped laser would have arrays of diodes wrapped around the amplifier material to provide intense flashes of light at a high repetition rate. NIF researchers have built a demonstrator laser called Mercury that Moses says can produce the required repetition rate but has a narrow aperture and hence low power. Such lasers “have come a huge distance over the last decade,” he says.

Livermore's proposal for a post-NIF fusion energy project, called LIFE (Laser Inertial Fusion Engine), would produce up to 500 megawatts of power. LIFE would be “part of a continuum” from NIF, Moses says. The whole laser system would “have the look and feel of NIF but operate at 10 hertz rather than hours.” One potential stumbling block is the high cost of laser diodes. “We could operate it now. We can show the performance but not the economics,” Moses says.

Others, however, say leaping straight from NIF to a power plant is the wrong way to go. NIF's and LIFE's lasers, they point out, have to be very high energy because they are doing two jobs: compressing the fuel capsule and heating it to spark ignition. Some researchers predict that if you separate those functions and build two lasers, each optimized for a different job, you would have a much more efficient system. The compressing driver laser would be about one-tenth of the energy of NIF's, and the heating laser less energetic still.

This approach, known as fast ignition fusion, is under investigation at several centers, most aggressively at Osaka University in Japan. About 5 years ago, a team there led by Azechi proved the principle of fast ignition by first compressing a hydrogen-filled capsule to 600 times the density of a normal liquid using Osaka's GEKKO-XII laser and then raising its temperature to about 10 million kelvin with a second laser. The experiment produced fusion reactions but not ignition. On the strength of that result, the team embarked on the Fast Ignition Realization Experiment (FIREX).

The first phase of FIREX aims to reach the ignition temperature of more than 60 million kelvin. Then the team will attempt ignition in phase two. To get there, Azechi and his team have been building a new heating laser. The key to heating is not high energy but high power in very short pulses. The FIREX heating laser will be able to produce pulses with petawatt (1015 watts) power lasting just 10 picoseconds (10−11 seconds). But the process has not been easy. “The power is so high, everything breaks,” says Azechi.

To keep the high-power pulses from destroying amplifiers, laser researchers use a technique called chirped pulse amplification: They take a short pulse, stretch it out in time using a diffraction grating, which reduces its peak power, pass it through an amplifier, and then use another grating to compress it back into a short pulse, now with higher power. “We need huge gratings with very high precision,” says Azechi. “It's a technical challenge that has taken almost 5 years.” The team has one of the laser's four beamlines working and will complete the rest in the next 1 or 2 years, Azechi says. But he adds that he's worried about funding for fusion in Japan. “With the technology, I'm very optimistic. Politically, it's difficult. There are so many projects in this field.”

In Europe, fusion researchers are betting on fast ignition to jump straight to a demonstration power-producing reactor called the High Power laser Energy Research (HiPER) facility. The project—a collaboration involving labs in 11 countries led by Dunne's Central Laser Facility—has already completed a design and is in the middle of a 3-year preparatory phase during which a consortium, funders, and industry involvement will all be put in place. The plan is to begin construction in mid-decade if all goes well at NIF.

Work to develop HiPER's lasers is going on all over Europe, and a test-bed fast ignition laser called PETAL is under construction near Bordeaux, France. “I feel a degree of pride in the U.K. and Europe,” says Dunne, who currently leads the project. “We're in a strong position with the technology and strategy. We're at the leading edge of this field.” Moses, however, thinks fast ignition isn't as close to viability as some believe. “LIFE uses low-risk technology. Fast ignition and other schemes are more futuristic, but they should be supported,” he says.

## All for one?

In the United States, meanwhile, the recent focus on NIF has largely diverted the government's attention from what comes next. (LIFE is an in-house LLNL project.) That is beginning to change. Moses says both Energy Secretary Steven Chu and one of his deputies, Steven Koonin, have visited the lab in recent months to learn about NIF and LIFE. Koonin has tasked the National Academy of Sciences to report on the next steps to achieve inertial fusion energy. “Our goal is to create a U.S. national effort, perhaps with international participation,” Moses says.

If success at NIF sets the field moving again, the question of cooperation is bound to arise: Does the world need three projects doing roughly the same thing? “We're interested in collaborations. How it develops will be very interesting,” says Moses. Dunne, who will soon leave the United Kingdom to become director of the LIFE project at Livermore, says he will work to merge the LIFE and HiPER teams. “It's the best way to achieve a result in the shortest possible time,” he says. “It's not about whether laser fusion will happen, but who will make it happen.”

10. Nanolasers

# Ever-Smaller Lasers Pave the Way for Data Highways Made of Light

1. Robert F. Service

New materials and techniques are bringing researchers close to a once-unthinkable goal: optical devices tiny enough to work hand in hand with electronic circuits.

The dream of optical computing—replacing electronic devices with much faster ones based on light—has tantalized scientists for generations. Nowadays, computer circuitry has grown too complex to be replaced wholesale. Instead, researchers talk about using lasers and other optical components as high-speed data highways between specialized electronic processors on chips. So far, lasers and other optical components have been far too big to make this integration possible. “It's hard to integrate the two technologies when the optical devices are 1000 times larger than the electronic devices,” says Cun-Zheng Ning, a physicist at Arizona State University, Tempe.

But the gap is narrowing. In recent years, researchers around the world have married traditional optical materials with metals to create lasers a mere tens of nanometers thick. Just last month, two groups reported making lasers ultrasmall in all three dimensions. “That was almost unthinkable before,” says Peidong Yang, a chemist at the University of California (UC), Berkeley. “There has been a lot of progress.”

Shrinking conventional lasers long seemed all but impossible. Traditional devices produce laser light by sending photons through an optical cavity made from a “gain” material and bouncing them back and forth with tiny mirrors. Along the way, the photons cause energized electrons in the gain material to release their energy as additional photons of light with the same wavelength. Those released photons then stimulate the release of still more photons. As the avalanche grows, some of the light is allowed to leak out of one of the mirrors to produce a tight beam of photons whose waves all travel in lockstep.

In most lasers, the gain region of the optical cavity is micrometers to meters in length. Photons of light can't be confined to spaces smaller than half their wavelength; otherwise they leak out. Making a smaller device lase is like pouring water into a sieve. “A classical [laser] cavity does not work,” Yang says. As a result, conventional lasers and other optical components for visible light can't be much smaller than 200 to 300 nanometers across, and most conventional optical components are much larger than that. Current transistors and other electronic devices, by contrast, include features that measure just tens of nanometers across.

The effort to shrink lasers took an important step in 2001, when Yang and his colleagues reported making the first laser from a nanowire, a whisker of zinc oxide as thin as 100 nanometers across (Science, 8 June 2001, p. 1897). Still, even though these lasers were ultrasmall in two dimensions, they were 4 to 10 micrometers long, so that photons traversing the optical cavity would have room to build up enough gain to lase.

Glimmers of a breakthrough came in the late 1980s and 1990s from researchers studying the way light interacts with electrons in metals. Metals are strong light absorbers. But researchers found that shining a light beam at the interface between a metal and a non-conductive (or “dielectric”) medium such as glass caused mobile electrons at the interface to oscillate back and forth at the same frequency as the light but with a much shorter wavelength. That discovery offered the hope that by coupling a laser's gain medium with a dielectric-metal interface, they could effectively squeeze the wavelength of the light produced by the device and thus make the overall device smaller.

Making such lasers took several years to pull off, but recent progress has been brisk. In 2007, researchers led by Martin Hill, an electrical engineer at Eindhoven University of Technology in the Netherlands, were the first to show that a semi-conductor surrounded by a thin metal structure could lase. More recently, Hill, Ning, and colleagues reported in the 18 June 2009 issue of Optics Express that they had made more complex lasers from pillars of indium-gallium-arsenide semiconductor as thin as 90 nanometers surrounded by a 20-nanometer-thick silicon nitride dielectric and a silver casing. Hill and his colleagues showed that the ultrasmall structure would lase when fed electricity—an impressive feat, as most new lasers need to be “pumped” with photons from another laser. Shaya Fainman, an electrical engineer at UC San Diego, calls the new work “pioneering” and says he's “a little jealous” of the paper. But he notes that the device works only at cryogenic temperatures. “To make it practical, it has to be electrically pumped and work at room temperature,” Fainman says.

Last month, Fainman's UCSD group reported that they had achieved this higher temperature operation. In their device, the UCSD researchers created an indium-gallium-arsenide-phosphide cylinder for the gain material. They then enveloped all but the bottom of the cylinder with a thin dielectric layer of silica and an outer layer of aluminum, leaving the bottom exposed to the air. During operation, the researchers fire a 1.064-micrometer laser beam at the open part of the pillar. The device uses that energy to create laser light at a longer 1.43 micrometers, which is then reemitted out of the uncoated bottom region of the pillar. Fainman says that they got the device to operate at room temperature by precisely engineering the width of the dielectric and metal layers. Now they must get the device to work when fed with electricity instead of a separate laser beam. “We are working on it,” Fainman says.

A very different-looking device was recently reported by physicist Xiang Zhang and colleagues at UC Berkeley. In the 1 October 2009 issue of Nature, Zhang's group reported creating a plasmonic laser using a long, thin cadmium sulfide (CdS) nanowire placed atop a 5-nanometer-thick dielectric layer of magnesium fluoride, which in turn rests on a silver surface. The device starts out much like Yang's original nanowire laser: First, the 10-micrometer-long CdS nanowire is pumped with a blast from a separate laser. Those photons are absorbed in the semiconductor, creating energetic electrons paired with electron vacancies called holes. When those electron-hole pairs recombine, they emit photons at a longer wavelength.

Some of those reemitted photons, Zhang explains, bounce back and forth in the nanowire as in a traditional nanowire laser. Most, however, strike the metal-dielectric interface and generate oscillating electrons known as plasmons, which can travel like ripples in a pond. Typically, such plasmons would travel only a very short distance in a metal before dissipating. But by sandwiching the magnesium fluoride dielectric between the semiconducting CdS nanowire and the metallic silver, Zhang says, the researchers can channel the plasmons into it. Because the dielectric material doesn't absorb plasmons, they can travel long distances back and forth. Tails of the plasmons' electromagnetic waves extend into the nanowire and prod it to emit more photons at the same plasmonic frequency, creating the hallmark avalanche effect of a laser. One end of the nanowire is slightly more leaky, and plasmons emerging from the structure produce a beam of greenish blue light.

In a paper posted on the arXiv preprint server* on 23 April, Zhang and colleagues report creating a related device that improves on the one they described in Nature. The new device works at room temperature instead of cryogenic temperatures, and it shrinks the longest dimension of the laser down to a single micrometer. Zhang says his team is now working on an electrically pumped version of their plasmonic laser.

Progress is coming from other angles as well. Last year, for example, researchers in Virginia, New York, and Indiana reported making a laser from 44-nanometer-sized particles (see sidebar). And last month, Yang's group at UC Berkeley reported having made an ultraviolet laser smaller than the wavelength of light in all three dimensions using a different technique known as “whispering gallery mode” lasing. Michael Haney, a physicist who manages research in this area for the Defense Advanced Research Projects Agency in Arlington, Virginia, says such progress shows that the goal of making lasers on the scale of electronic devices is “advancing very quickly.” The next challenges, Haney says, are to get usable light out, make the lasers electrically activated, and assemble them in large arrays that can be integrated with electrical components. Those hurdles won't be easy to clear. But the progress achieved thus far makes the prospects much brighter.

11. Nanolasers

# Smallest of the Small?

1. Robert F. Service

Lasers have been shrinking ever since they were invented 50 years ago. But the 44-nanometer "spaser" will be hard to beat.

Lasers have been shrinking ever since they were invented 50 years ago. But the 44-nanometer “spaser” will be hard to beat.

The device, reported online in Nature on 16 August 2009 by researchers at Norfolk State University in Virginia; Purdue University in West Lafayette, Indiana; and Cornell University, is akin to a conventional laser. In traditional lasers, light bounces between two mirrors, stimulating a “gain” medium to release more photons of light at the same wavelength to produce a tight, concentrated beam. In their surface-plasmon laser, or spaser, Mikhail Noginov, a physicist at Norfolk State, and his colleagues shine a conventional bluish-green laser beam into a suspension of nanoparticles, each with a gold core surrounded by a layer of sodium silicate and an outer silica shell containing dye molecules. When the gold is excited by the laser photons, its surface starts to dance with collective oscillations of electrons, known as surface plasmons. The plasmons tickle the dye molecules to release more plasmons at the same wavelength, causing the device to emit green laser light.

“It's impressive work,” says Cun-Zheng Ning, a nanolaser expert at Arizona State University, Tempe. But other outsiders are more cautious. “There are questions,” says one physicist, who asked not to be identified. Among the issues is a concern that the lasing could be a collective response from clusters of nanoparticles rather than from individual particles. Noginov and his colleagues respond that aggregation of particles, which can potentially affect their stimulated emissions, would also cause a shift of the surface plasmon spectrum. “This was not the case of our experiment,” Noginov says.

Noginov says efforts to use this spaser design in high-speed computing could confront a greater challenge. The organic dye molecules used in the nanoparticles are fragile. To make a device robust enough for use in computing technology, Noginov says, researchers will likely need to make new all-inorganic-based spaser nanoparticles, a goal the field now faces.

12. Optics

# Putting Light's Light Touch to Work As Optics Meets Mechanics

Forces exerted by light can set tiny objects aquiver, a phenomenon scientists hope to harness in the burgeoning field of cavity optomechanics.

The effect appeared unexpectedly. In 2005, Kerry Vahala, an applied physicist at the California Institute of Technology (Caltech) in Pasadena, and his team were experimenting with microtoroids—little glass disks that resemble dinner plates with fat, rounded rims. Light can circulate around the rim, and each microtoroid resonates with laser light of a specific wavelength and frequency, just as an organ pipe rings at a distinct pitch. But the researchers found that the light passing through such a doohickey also warbled up and down in intensity at a much lower frequency (Science, 15 July 2005, p. 366).

Eventually, the researchers found the source of the radio-frequency warbling: Pressure from the light within the toroid was making it vibrate. “We were not looking for this [mechanical effect] at all,” Vahala says. “There was no precedent for it aside from some theoretical speculation.”

That light-and-motion connection has spawned a new field. Scientists have long used laser light to move atoms and molecules—Nobel laureate and U.S. Secretary of Energy Steven Chu developed “optical tweezers” that can uncoil DNA. But researchers are now using light to control the motion of larger humanmade objects built of materials like glass or silicon. A device must be both a fine “optical cavity” that rings with light and an outstanding “mechanical resonator” that, like a pitchfork, vibrates readily at a precise frequency. Researchers can then use light to control vibrations or vice versa.

Such “cavity optomechanics” might be used to coax vibrating machines to hum to the odd rules of quantum mechanics, which ordinarily govern the realm of atoms and molecules. More practically, optomechanical widgets might process optical signals or serve as frequency standards. One team has even used them to make a kind of laser for vibrations. Other exciting results are sure to come, says Oskar Painter, an applied physicist at Caltech: “This wave hasn't crested yet.”

## All aquiver

Light can shake an object in several ways. In Vahala's gizmo, light's pressure does the trick. A toroid will resonate with light bled in from an optical fiber if its circumference equals a multiple of the light's wavelength. As light accumulates, its pressure stretches the toroid (see figure, below). But that stretch spoils the resonance, so the circulating light wanes, the pressure eases, and the toroid shrinks. The cycle of expansion and contraction repeats at a rate set by the stiffness of the disk, typically millions of cycles per second, and makes the light leaving the cavity warble.

Light exerts a sideways “gradient force” as an insulating material feels a pull to where the light is most intense, like an actor who craves the spotlight. For example, if one toroid lies atop another, then gradients in the light leaking between them can pull their edges together or push them apart, as Michal Lipson, an applied physicist at Cornell University, and colleagues reported online 15 November 2009 in Nature. If the toroids are springy, then that force can also trigger their edges to flap up and down in a cycle akin to the one pressure causes in a single toroid, as Painter reported on 31 August 2009 in Physical Review Letters.

Light can also trigger vibration by making a material contract. Tal Carmon, an applied physicist at the University of Michigan, Ann Arbor, and a colleague circulated light in a 100-micrometer-wide glass sphere. In a chicken-or-egg process called stimulated Brillouin scattering, the light set off a vibration moving in the same direction, while the vibration “scattered” light into a second wave of a longer wavelength going the other way. The overlapping light waves created a moving pattern of bright spots that caused the glass to contract and amplified the vibration, as the team reported on 19 March 2009 in Physical Review Letters.

Light is a wimpy mover and shaker; a laser beam with a power of 1 watt exerts a force of just 6 nanonewtons. To respond to such feeble shoving, an object must resonate with light, so that each photon circulates in it thousands of times, greatly enhancing the photon's mechanical effect. The device must also ring with vibrations, so a small force repeatedly applied can cause a large movement. Only recently has it been possible to combine those properties. “These structures weren't around 15 years ago,” Vahala says.

## Cool it!

Now that the gadgets exist, physicists want to see what they can do. Many are trying to coax them to wiggle quantum mechanically. According to quantum theory, a vibrating object, or “oscillator,” can absorb energy only in discrete “quanta” whose size is set by the vibration frequency. The object can't stand still, either; it will always possess an inextricable half-quantum of energy and hum with “zero-point motion.” Physicists hope to spot that unquenchable quaking as a step toward more-bizarre states of motion, such as one in which an oscillator is in two places at once.

Each quantum is minuscule. So to remove every possible one and reach the “ground state,” physicists must cool an oscillator to nearly absolute zero. Ironically, Vahala's observation that light can trigger vibrations opens the way to cooling an oscillator by shining a laser on it. The trick is to reverse the process, says Tobias Kippenberg of the Swiss Federal Institute of Technology, Lausanne. “The fact that you could use this for cooling was immediately clear,” he says.

To trigger a vibration in, say, a toroid, the light's frequency must be tuned slightly higher than the toroid's resonant frequency. To still a vibration, the light's energy and frequency must be lowered—optimally, by the frequency of the vibration. To enter the toroid, a photon must then make up the energy it lacks by absorbing a quantum of energy from the toroid. The scheme works because, thermodynamically, “a laser beam is a very cold object,” says Jack Harris of Yale University. “A good laser is at 0 kelvin.”

Last year, three groups used laser cooling to reduce an oscillator's energy to a few dozen quanta. However, they were beaten to the ground state by a team that simply chilled an oscillator with a very high frequency—and hence large quanta—in a liquid-helium refrigerator (Science, 29 January, p. 516). But optical methods still have advantages, says Markus Aspelmeyer of the University of Vienna. “That's why the result has only encouraged us,” he says.

Laser cooling alone might chill an oscillator to the ground state, Aspelmeyer says. That might make it possible to run an array of oscillators not deep within a helium refrigerator but on a more-accessible optical table to process information encoded in the quantum state of light. And optical methods might do things that other methods can't, Harris says. For example, he has devised an optical scheme to watch the quantum jumps when an oscillator absorbs individual quanta. Optomechanical devices could also generate exotic quantum states of light, Harris says, such as “squeezed” states that suppress the random lumpiness of the photons in the light.

## Other vibes

Optomechanical gizmos might serve myriad other purposes, too. “I'm very excited about the applications of these structures,” says Cornell's Lipson. A double-toroid makes a low-power optical switch, she says, as one laser beam can change the size and optical frequency of the thing and control the passage of another beam. A jiggling toroid makes a frequency standard that's immune to electrical interference. It works as a “down-converter to extract a radio signal encoded in an optical carrier beam, as Mani Hossein-Zadeh, an applied physicist at the University of New Mexico, Albuquerque, reported on 15 February 2008 in IEEE Photonics Technology Letters.

Meanwhile, other researchers are working on more speculative ideas. In the 1990s, physicists pioneered “photonic crystals,” materials riddled with holes through which photons travel like the electrons in a crystalline solid, with only certain energies allowed. Similar holey “phononic crystals” control vibrations in the same way. Now, Caltech's Painter and his colleagues have combined the two, as they reported online 18 October 2009 in Nature.

Their optomechanical crystal consists of a ladderlike beam of silicon about 30 micrometers long and 1.3 micrometers wide (see figure, right). The spacing between rungs shrinks slightly near the beam's middle, creating a “defect” that traps both light of a certain frequency and vibrations shaking at about 2 billion cycles per second. Again, light forces set off vibrations, which reveal themselves in a warbling of the light.

Next, the team used optical gradient forces to couple the jiggling of two parallel bridges, tuning their combined motion to a so-called dark state that neither emits nor absorbs light, as they reported 7 February in Nature Photonics. That feat raises the possibility of snaring a light pulse in that dark state. “We've got exactly what we want, which is the ability to store a photon pulse and then release it,” Painter says. Physicists had previously stopped light in atomic vapors, but optomechanical crystals are easier to control and could be used for all-optical quantum information processing, he says.

Physicists have even fashioned the equivalent of a laser for vibrations. In a laser, a light-emitting material sits between two mirrors. The atoms in it are boosted from their ground state to a more energetic, excited state by, say, applying a voltage, and as they “de-excite,” the atoms emit photons, which bounce between the mirrors. Passing through the material, the photons trigger excited atoms to emit more photons of the same wavelength and direction. Such “stimulated emission” produces a torrent of identical photons that is the laser beam.

Twisting that idea, Caltech's Vahala amps up vibrations using jumps between optical states of two toroids placed side by side. As light bleeds from one to the other, their resonances meld into a lower frequency one and a higher frequency one, just as the electronic orbitals of two identical atoms meld and split when the atoms form a molecule. Adjusting the toroids' spacing, the researchers tune the frequency difference to equal the frequency at which one toroid shakes, as they reported 22 February in Physical Review Letters. When they then pump in enough high-frequency light, the vibrations amplify themselves by stimulating jumps to the low-frequency optical state.

Where all these efforts will lead remains to be seen. Some researchers say that, technologically, optomechanical devices have yet to do anything that electronics can't do. “I don't see anything that's a breakthrough at this point,” Hossein-Zadeh says. However, the tiny machines may find uses in basic research in various fields of physics. “These little widgets will be a part of science, even outside of what we call cavity optomechanics,” Vahala says. Now that physicists can shake things up with light, likely they'll find ways to use that trick.

13. # Redrawing Africa's Malaria Map

1. Martin Enserink

Smaller countries on the hardest-hit continent have shown how aggressive malaria control can slash cases and deaths. Can the big ones follow?

Africa is the key battlefield in the fight against malaria—and in recent years, there have been major successes. Thanks to the widespread introduction of insecticide-treated bed nets, indoor spraying, and a new generation of drugs called artemisinin-based combination therapies (ACTs), a dozen countries have dramatically reduced their disease and death toll faster than anyone expected. “The number of cases and deaths has fallen shockingly rapidly,” says Robert Newman, head of the World Health Organization's Global Malaria Programme.

But most of the successful countries were small and peaceful. More than half of all malaria cases occur in Nigeria, the Democratic Republic of the Congo, Sudan, Tanzania, and Uganda—and in some of those, the challenges are far greater. Other questions loom as well. Can the countries that have seen malaria dwindle keep up their vigilance? And will donors remain interested?

Looking at the map of Africa, however, malaria fighters say they can't help but feel a little relieved. “There's still a lot to do,” says Richard Steketee of the malaria control program at PATH in Ferney-Voltaire, France. “But we're in a very different place than we were 10 years ago.”

14. # As Challenges Change, So Does Science

1. Martin Enserink

Although nobody believes that it is possible in the next 10, 20, or even 30 years, the call to eradicate malaria, combined with the plummeting disease burden, are already reshaping the scientific agenda.

Malaria scientists are brimming with optimism. Gone is the gloomy atmosphere that dominated the '90s, when nothing seemed to get done and funding was always in short supply. Thanks to a massive influx of new money, dozens of countries have begun scaling up simple ways to fight malaria: insecticide-treated bed nets, indoor spraying of insecticides, and a class of powerful drugs called artemisinin-based combination therapies (ACTs). Some have seen spectacular results. On top of that, megaphilanthropists Bill and Melinda Gates gave the field a breathtakingly ambitious goal in 2007: Don't just control malaria, eradicate it—get rid of it once and for all.

Nobody believes that is possible in the next 10, 20, or even 30 years. But the call for eradication, combined with the plummeting malaria burden, are already reshaping the scientific agenda. Rather than preventing disease and death, there's a new focus on breaking the chain of transmission between host and parasite. Countries that have brought down cases by more than 90% are asking, “Now what?” Can they eliminate malaria—that is, get local transmission down to zero? The number of papers mentioning eradication and elimination in the context of malaria has exploded.

Over the past 18 months, a group of 15 scientists, supported by the Gates Foundation and led by Pedro Alonso of the Barcelona Centre for Inter national Health Research in Spain, has consulted with hundreds of experts to hash out a new road map, called the Malaria Eradication Research Agenda (malERA). The fruits of their labor will be published in the months ahead. But based on interviews with malaria scientists and previous papers, here's a look at what's up and down on the research agenda.

Transmission-blocking drugs Experience with ACTs has shown that drugs don't just save individual patients. They can also prevent others from getting sick, because treated patients become less infective to mosquitoes that bite them. But ACTs aren't ideal for this purpose, because they don't kill all the gametocytes, the cells in the parasite's sexual stage that get picked up by the mosquito. Another drug, primaquine, does a much better job, but it also causes the breakdown of red blood cells in patients with certain variants of a gene called G6DP. So safer gametocyte-killing drugs are needed, as well as research into the possibility of screening patients for their individual risk of primaquine side effects.

Partially effective vaccines Developing a fully protective malaria vaccine has proven elusive. But because of the disease's terrible burden, researchers have long believed that even a partially effective vaccine can do a great deal of good, and one such vaccine—GlaxoSmithKline's RTS,S, which offers about 50% protection the first 8 months—is now in a phase III trial in seven African countries.

But the surprising successes achieved by existing cheap tools make these vaccines less interesting. “Maybe we won't need a vaccine that does what bed nets can do,” says Carlos Campbell, who directs the malaria control program at PATH, a Seattle, Washington–based nonprofit.

Transmission-blocking vaccines The hunt is on for new generations of vaccines that don't protect the recipient from being infected but prevent mosquitoes from passing on the parasite to other people. Several of them are in development (see main text, p. 847).

Intermittent preventive treatment In so-called intermittent preventive treatment (IPT), malaria drugs are given to people at high risk, such as children and pregnant women, in malaria-endemic areas to prevent them from becoming sick. That strategy becomes less attractive when malaria cases have been sharply reduced, because it means treating more and more healthy people for one malaria case prevented. Indeed, some countries where malaria rates have plummeted are beginning to question the use of IPT in pregnant women. Instead, research is needed to find other ways to protect high-risk groups, scientists say.

Mass drug administration Treating entire populations with malaria drugs has been tried since the 1930s; chloroquine or amodiaquine was even added to salt in a dozen countries. The latter strategy is now widely seen as “stupid,” says Brian Greenwood of the London School of Hygiene and Tropical Medicine; such massive use easily triggers resistance, and the spectacular successes in controlling malaria it brought were usually short-lived.

Today, however, scientists say a few rounds of carefully planned mass drug administration, combined with other measures, might deal parasites the coup de grâce in countries on the brink of elimination. But it will take a new drug combination that is safe enough to give to hundreds of thousands of people and that, ideally, is so effective that a single dose will do. Another option would be to screen entire populations and treat only carriers of malaria. Modeling studies are needed to predict the effects of both strategies.

Plasmodium vivax Plasmodium falciparum is a global mass murderer. Its lesser-known cousin, P. vivax, found in Asia, South America, and Africa, is far less deadly. From the eradication viewpoint, however, P. vivax is a bigger problem because it can lurk in the human liver for years and come back unexpectedly. P. vivax needs more study, researchers say—and new drugs are needed that kill the sleeper cells, called hypnozoites, in the liver. The Medicines for Malaria Venture has one called tafenoquine in a phase II trial.

15. # Malaria's Drug Miracle in Danger

1. Martin Enserink

Like many others before it, the latest generation of malaria drugs is losing its punch. This time, can global disaster be averted?

BANGKOK—Malaria rates are plummeting in many places, and scientists are optimistically talking about ridding entire countries of the disease—or even, in the long run, eradicating it worldwide. But in Southeast Asia, a new threat is looming—one that so far has received little attention but that could wipe out the recent advances and set back the global fight by decades.

Along the border between Thailand and Cambodia, Plasmodium falciparum, the most dangerous of malaria parasites, is showing unmistakable signs of becoming resistant to artemisinin derivatives, the group of powerful drugs that—as part of so-called artemisinin-based combination therapies (ACTs)—have become the cornerstone of malaria treatment worldwide.

For the moment, the drugs still work in the area; they just take more time to do the job. But that is alarming enough, scientists say. If resistance gets worse and starts spreading out from here, the results could be catastrophic. There are few other drugs in the pipeline, and those that are closest to the market may not work against the resistant strain (see sidebar, p. 846). One infected person flying to, say, India or Africa could unleash the resistant parasites there.

Plans are under way to minimize the risk of that happening. The idea is to control malaria as aggressively as possible in the border area, making sure patients don't pass resistant parasites on to mosquitoes, and keeping parasites from being exposed to nonlethal drug levels that could fuel more resistance. Already, large numbers of insecticide-treated bed nets have been distributed in the area, and Cambodia has enacted a ban on the sale of so-called monotherapies, pills that contain artemisinin but lack the double whammy provided by the second drug in an ACT.

The World Health Organization (WHO), which coordinates the containment effort, bills it as unprecedented; when previous malaria drugs started becoming ineffective, the world sat by and watched it happen, says Robert Newman, head of WHO's Global Malaria Programme. But some malaria experts aren't satisfied. “There have been endless meetings but not enough action,” says Nicholas White, who's widely considered the dean of malaria research in Asia. Because millions of lives are at stake, White, employed by the British Wellcome Trust and the University of Oxford but stationed at Mahidol University in Bangkok since 1981, says the best strategy would be to eliminate malaria from the border area altogether. That idea, discussed widely a few years ago, has so far gone nowhere.

Malaria scientists who talked to Science were careful not to come down too hard on WHO, because the agency is chronically understaffed and because they want to see Newman—a respected malaria veteran who's been on the job less than a year—succeed. But they say the agency needs to step up its efforts, and the world should pay more attention. “Blame is not relevant, but outrage is entirely appropriate,” says malariologist Christopher Plowe of the University of Maryland (UMD) School of Medicine in Baltimore. Given the huge stakes, White argues that the fight against resistence should “be prosecuted like a war. Because it is a war.”

## Déjà vu

If, a decade ago, you had asked malaria researchers where artemisinin resistance might first crop up, they probably would have put their money on the Thai-Cambodian border, a region that has been the cradle of resistance multiple times. In 1957, chloroquine, discovered by Bayer in the 1930s and deployed widely after World War II, began failing here. Resistant P. falciparum started spreading in all directions in Asia, as well as to east and eventually west Africa. A separate resistant strain that arose along the Panamanian-Colombian border fanned out across South America, and by 2000, the drug was virtually useless in most parts of the world.

Chloroquine's main successor, sulfadoxinepyrimethamine (SP), also known as Fansidar, started losing its potency here in the 1960s; this time resistance spread much faster. SP's replacement, mefloquine—developed by the U.S. Army during the Vietnam War and notorious among Western travelers for its side effects—started failing in Southeast Asia in the early 1980s. It's still effective in most other parts of the world.

Why the region is such a hotbed isn't clear, and the reasons are probably diverse. Widespread overuse of drugs may have played a role, as may the trade in counterfeit drugs. Fake drugs often contain lower levels of a drug—enough to make the patient feel better but too little to kill the parasites. (The same can happen when patients take low doses of a legitimate drug.) Some scientists have also speculated that the parasite population in this region may be more genetically diverse than elsewhere, giving it an evolutionary leg up. Whatever the causes, history appears to be repeating itself, sooner than anyone anticipated.

## Hard to track

Chinese scientists isolated artemisinin in the 1970s from sweet wormwood (Artemisia annua), a plant used to treat fever for centuries. The drug, along with more potent derivatives such as artemether, artesunate, and dihydroartemisinin, not only kills parasites effectively and fast but also disappears from the bloodstream in less than a day after administration. The hope was that this would give the parasite less chance to adapt and develop resistance, and that teaming each of the drugs with another, longer-lasting one that mops up any remaining parasites would further reduce chances of resistance.

Perhaps that's why many didn't believe there was a serious problem when the Thai Ministry of Public Health first reported that treatment with the combination artesunate-mefloquine—the most-used ACT in Southeast Asia—occasionally failed in the province of Trat, which borders Cambodia, in 2003 and 2004. (Thailand is almost entirely malaria-free; the disease occurs only along its borders.) Most experts believed the problem was resistance against mefloquine, the partner drug, a well-known problem.

Fingering artemisinin wasn't easy, says Harald Noedl of the Medical University of Vienna, who carried out the pioneering resistance studies with researchers from the U.S. Armed Forces Research Institute of Medical Sciences in Bangkok. Scientists haven't yet found genetic markers by which they can easily identify resistant P. falciparum strains. The only surefire way to track artemisinin resistance is to treat patients with the drug—and without the partner drug—and check how fast they clear the parasites. WHO and ethical panels have concluded that this is ethically acceptable, says Noedl, provided subjects get a 7-day course of artesunate—instead of the standard 3 days for the combination—and they are followed carefully for up to 9 weeks. Such “treatment trials” are time-consuming and expensive, and they have been slow to get off the ground.

One study in western Thailand showed slightly increased parasite clearance times, but not nearly as dramatic as in Cambodia. It's impossible to say whether the change represents a very early stage of resistance; patients could also take longer to get rid of malaria because malaria control has been so effective that their immunity has waned, says François Nosten, who heads the Shoklo Malaria Research Unit in Mae Sot, a town on the Thai border with Myanmar. If resistance spills over into Myanmar, where the health infrastructure is poor and malaria is rife, most experts agree it would be impossible to stop it from spreading farther west to Bangladesh and India.

Action is needed, and the Thai and Cambodian governments, WHO, and a range of donors have all stepped up their efforts; the Bill and Melinda Gates Foundation alone has contributed 22 million. What's been lacking, however, is a clear plan, a common strategy, and strong leadership, says White. “I'm in the epicenter here, and even I have trouble finding out what's happening,” he says. When Science asked WHO, which coordinates the efforts, about the current strategy, Newman said WHO has a plan but conceded that no publicly available version of it existed; he had WHO staff members write up a strategy and a progress report, however, which he sent to Science on 28 April. Among the accomplishments listed in the report: Cambodia is improving its surveillance and has recruited more than 1300 volunteer malaria workers to help diagnose patients early and give them the right treatment. Bed nets are ubiquitous, and malaria rates are dropping. Cambodia also banned monotherapies in 2009, and that seems to be working, at least in some places, says Arjen Dondorp, the head of malaria research in White's group; when Dondorp asked a Cambodian co-worker last year to see whether he could buy the drugs at local pharmacies in the province of Pailin, the man came back empty-handed. But the challenges remain formidable. Western Cambodia has many migrants from neighboring countries looking for work, and they're difficult to find and treat. Nobody even knows how many people are in the area, says Charles Delacollette, the coordinator of WHO's Mekong Malaria Programme. Bed nets don't work as well here as in Africa because mosquitoes start biting earlier in the evening, before people go to bed, and bite outside homes as well. The tensions between Thailand and Cambodia—just a few weeks ago, their armies briefly exchanged shots in a disputed border zone—don't help. One urgent question is whether a new, so-called synthetic artemisinin might work against resistant parasites. A study to find out was approved by ethical panels and ready to go when Cambodia's minister of health unexpectedly vetoed it; apparently part of the reason was that researchers from Thailand were involved, says Stephan Duparc, chief scientific officer of the Medicines for Malaria Venture, which supports the study. Still, experts from both sides of the border do meet frequently without political interference, says Wichai Satimai, director of Thailand's Bureau of Vector-Borne Disease. The idea of wiping out malaria from the region altogether, meanwhile, has for now been shelved. On paper at least, the plan has potential. In terms of malaria, the area is almost an island, surrounded by non-malarious areas to the north, west, and east and by the Gulf of Thailand to the south. If the disease were eliminated, it might bounce back—but it would do so by being reimported from another region, such as eastern Cambodia, where there is no resistance. White believes an approach called mass drug administration, in which everybody in the area receives drugs in a short time (see sidebar, p. 843), could do the job. “It would be a massive operation. You'd need global positioning systems, the military, and who knows what else to find all the people,” he says, “but I think it could be done.” But WHO is generally not in favor of mass drug administration, which targets people regardless of whether they have malaria and carries the risk of aggravating resistance if it's not carried out properly. WHO describes elimination as the “ultimate goal.” Newman resents the perception that the threat is not being taken seriously enough. WHO has lobbied hard against monotherapies, he points out, and WHO Director-General Margaret Chan has raised the issue of artemisinin resistance multiple times. But UMD's Plowe says that isn't nearly enough. “There should be a czar for malaria elimination in Southeast Asia, with a strong international team,” he says. “Right now, I don't know anyone whose full-time job it is to contain resistance.” White agrees; the stakes are high enough, he says. “I would hate for resistance to show up in Mali next year and to know that we didn't stop it. I think we would all feel that we had failed terribly.” 16. # If Artemisinin Drugs Fail, What's Plan B? 1. Martin Enserink If the current generation of artemisinin-based combination therapies should fail, does the world have a solid backup plan? The short answer: No. If the current generation of artemisinin-based combination therapies (ACTs) should fail, does the world have a solid backup plan? The short answer: No. New funding has lured academics and pharmaceutical companies back to the malaria field after a decades-long drought. But drug development is a slow process, and the only compounds far enough along in the pipeline to quickly replace current ACTs are variations on the artemisinin theme. Whether they will kill resistant parasites is a big question. The three artemisinin derivatives currently used in ACTs are so closely related chemically that parasites resistant to one will probably be resistant to all, says Harald Noedl of the Medical University of Vienna. In other words, you can't swap one for the other. One possible alternative is a new artemisinin derivative, developed by the Hong Kong University of Science and Technology, whose structure is different from that of the others. The drug, called artemisone, has been through a phase II trial for nonresistant malaria but didn't offer major benefits over existing derivatives, says Timothy Wells, chief scientific officer at the Medicines for Malaria Venture (MMV), a nonprofit that links scientists and companies to get new drugs to the market. It might have a new future, however, if the resistant parasites found along the Thai-Cambodian border prove sensitive to it. MMV is supporting a proposed trial in eastern Thailand to find out. Then there are the synthetic artemisinins. These drugs, developed by researchers at the University of Nebraska, Monash University in Australia, and the Swiss Tropical Institute, were originally designed to circumvent the shortages and high prices of sweet wormwood, the plant from which artemisinin is isolated. Made from scratch in the lab, the molecules have the key chemical feature of artemisinin, a so-called endoperoxide bridge, but other than that, they are completely different. MMV stopped sponsoring development of the first of these compounds, arterolane, after a disappointing phase II trial for nonresistant malaria, but it has renewed its collaboration with Indian pharma giant Ranbaxy to test it against resistant malaria, paired with a drug called piperaquine in an ACT. The three academic groups, meanwhile, have developed a second, more potent candidate, provisionally called OZ439, that will also be tried as soon as it has proved to be effective against nonresistant malaria in a phase II study. Based on their different chemical structures, Wells feels chances are “pretty good” that the synthetics can replace standard artemisinins if they fail. Others are less sanguine because the mechanism of resistance is still unknown. There's a chance of success if, for instance, parasites become resistant through a pump that removes the drug from their cells, says Arjen Dondorp of Mahidol University in Bangkok; such pumps are unlikely to work for two molecules with distinct shapes. But if resistance occurs because the drug's target has changed a bit—making the fit less snug—then the new synthetics may not do any better. Beyond artemisone, arterolane, and OZ439, there's not much else in sight. There are plenty of new leads, but none of them are even in phase I safety studies, Wells says, and they would take at least 7 or 8 years to get to the market. 17. # The ‘Do Unto Others’ Malaria Vaccine 1. Gretchen Vogel Once neglected, research on transmission-blocking vaccines for malaria is gaining new prominence. It used to sound like a far-out idea: a malaria vaccine that would use humans to generate antibodies and deliver them to mosquitoes, with the aim of preventing the insects from spreading the disease. Until recently, such transmission-blocking vaccines (TBVs) had received scant attention. But after Bill and Melinda Gates called for a global effort to eradicate malaria (Science, 7 December 2007, p. 1544), TBVs have moved to center stage in malaria vaccine research (see p. 843). With money no longer the limiting factor, progress is accelerating: A half-dozen TBVs could have a shot at clinical trials sometime in the next 5 years, researchers say. Even so, the scientific and social hurdles remain daunting. Key among them is whether people can be persuaded to get a vaccination that doesn't prevent them from getting sick but instead protects family and neighbors from getting infected. That also raises the bar for an extremely safe vaccine. Nirbhay Kumar, a molecular parasitologist at Tulane University in New Orleans, Louisiana, has been working to develop a TBV for 28 years, often with little support. But he understands why few funders shared his enthusiasm: With limited resources, it made sense to focus on fighting the disease in humans. Consequently, the few candidate vaccines developed to date attack the parasite during the human stages of its life cycle. The frontrunner, called RTS,S, targets the parasite in the liver stage—when there are hundreds of parasites (see graphic). Other candidates target the blood stage, when there can be trillions of parasites. So far, such vaccines protect people from symptoms of malaria but don't clear all parasites from the blood. As a result, a vaccinated person could be symptom-free but still be infected and infect others. “You can't eliminate or eradicate [malaria] without addressing the transmission” of the disease, says Kumar. “If we want to completely finish the job, we will need that vaccine,” says Christian Loucq, head of the PATH Malaria Vaccine Initiative (MVI), a nonprofit funded primarily by the Gates Foundation. ## Playing the numbers Antibodies generated by TBVs would be transferred to a mosquito during a blood meal, and they would zero in on the parasite after it enters the insect. That gives them a built-in advantage, says parasitologist Robert Sinden of Imperial College London: Instead of mopping up thousands or even trillions of parasites at once, they have to take out only a handful. The disadvantage? The vaccinated person can still get sick. But a mosquito that bites him or her won't pass the disease to others. But even with the numbers advantage, developing this new class of vaccines isn't easy. TBV researchers not only have to aim at entirely new targets but they must also overcome many of the same problems that have plagued other malaria vaccines. One of those problems is that Plasmodium proteins—the targets of any type of malaria vaccine—are difficult to make in the lab. Usually vaccine producers use bacterial, yeast, or animal cells either to grow the target microbe directly or to produce recombinant proteins, but the malaria parasite's genome can confound the proteinmaking machinery in these cells. To make matters worse, the antigens must have the right shape to trigger effective antibodies, and getting the recombinant Plasmodium proteins to fold correctly has proved difficult. Moreover, because TBVs won't provide direct benefit to the recipient, some say, any TBV with a chance in the real world has to be flawless: absolutely safe, 100% effective, and cheap. A few teams began testing TBVs in humans a decade ago. The first antigen out of the gate was called Pfs25, named for a 25-kilodalton protein on the Plasmodium falciparum ookinete, the fertilized zygote that migrates through the mosquito midgut. Scientists have also tested the analogous protein from P. vivax, called Pvs25. (Researchers call both proteins P25 for short.) Like other Plasmodium proteins, P25 often failed to fold correctly when expressed in recombinant cells, but animal tests were promising enough to try it in humans. In initial clinical trials, vaccines based on P25 prompted measurable but weak reactions in healthy volunteers. Based on that experience, malaria experts Yimin Wu and Louis Miller of the National Institute of Allergy and Infectious Diseases in Rockville, Maryland, and their colleagues began a trial in 2005 that combined the antigens with an adjuvant that boosts the body's immune response to a vaccination. The researchers planned to test the vaccine in 72 healthy volunteers. Initial results showed that some recipients produced antibodies that could block 90% of oocyst development in mosquitoes. But severe reactions in two participants stopped the trial. The combination of adjuvant and antigen was apparently too much for some immune systems, says Anna Durbin, a vaccine expert at the Johns Hopkins Bloomburg School of Public Health in Baltimore, Maryland, who coordinated the trial. Several groups are continuing work on P25 and the related P28, however, hoping that better folding might produce a potent response without new adjuvants. At the same time, ongoing research on a half-dozen other candidates has brought them within range of clinical trials. One is Pfs48/45, about twice as big as P25 (and therefore harder to produce), which is expressed during the sexual stages of the parasite, both in the human and in the mosquito. Researchers identified it as a potential target and cloned the gene in 1993, “and we spent the next 15 years on its folding,” says parasitologist Robert Sauerwein of Radboud University Nijmegen Medical Centre in the Netherlands. Eventually, several groups developed tricks to produce the protein in its lifelike shape. Kumar and his colleagues have conducted promising trials in baboons: The blood of animals that received a single immunization blocked parasite development by 93%. Another candidate, called P230, is even more of a “monster,” says Christine Farrance of the Fraunhofer USA Center for Molecular Biotechnology in Newark, Delaware. It is also expressed during the sexual-stage parasite, but it is 230 kilodaltons, too big to make in cell culture. Farrance and her colleagues are working with several teams to produce parts of P230 in tobacco plants, which are an inexpensive way to make complex proteins. Another leading candidate, called HAP2, is sometimes called the coitus interruptus antigen. First discovered in plants, but also present in Plasmodium, it is crucial for the fusion of male and female gametes. In lab tests, Sinden says, it can block parasite development by 96%. One of the newest targets is particularly promising. Instead of targeting a parasite protein, the candidate vaccine targets a protein in the mosquito that is the docking point for the parasite as it invades the insect's gut. Presumably, the antibodies bind to the gut protein, blocking the parasite's way and preventing its further development. One advantage to targeting a mosquito protein is that the antibodies are effective against several strains of para site, says Rhoel Dinglasan, a malariologist at Johns Hopkins Malaria Research Institute in Baltimore. Antibodies from rabbits vaccinated with the antigen, called AnAPN1, blocked development of P. vivax by 98% in mosquitoes in Thailand and completely blocked P. falciparum development in mosquitoes in Cameroon, Dinglasan says. “It's the first time a candidate vaccine could do two with one blow,” he says. The ability to hit both parasites with a single vaccine is a major advantage, says Ashley Birkett of MVI, which is collaborating with Dinglasan in one of its first forays into TBV research. ## Clinical conundrum As the various antigens move closer to human testing, experts are pondering how to design clinical trials. Vaccine clinical trials are never easy, but proving that a TBV can be effective in real life will pose new challenges, because the protection they offer is to communities, not to the vaccinated individual. Scientists have only begun to think about these issues, says epidemiologist Thomas Smith of the Swiss Tropical and Public Health Institute in Basel. “Large-scale trials of TBVs did not seem to be an imminent possibility until very recently,” he says. And because recipients don't benefit directly, regulatory agencies may require even more safety data before approving them. The epidemiology of malaria varies widely, and so vaccination strategies will differ from place to place. But the most likely scenario for an initial real-life test, Smith says, would involve several dozen communities with relatively low transmission. In half the communities, researchers would vaccinate as many people as possible with the TBV. The remaining villages would serve as controls. “The readout would be whether transmission was interrupted in the vaccinated communities,” Smith says. To get a handle on the factors that could affect a trial, Loucq says, MVI will hold a workshop in June that will bring together regulatory agencies, experts in mosquito behavior, epidemiologists, and statisticians who can help assess variables such as how far mosquitoes travel and how often they take a blood meal. If a TBV proves effective in the field, the next step will be to see whether it could be combined with a vaccine that would prevent disease. “A multicomponent vaccine is what everyone is thinking about,” Dinglasan says. If a transmission-blocking component could be added to RTS,S, Birkett says, “you would get the best of both worlds: a vaccine that has direct benefit to the recipient but also provides the community protection.” At the same time, he says, any malaria vaccine has to be “very affordable, and combining antigens is much easier said than done.” A TBV will become increasingly important as other malaria-control measures have an effect and transmission falls, Sauerwein says. “It's very difficult to find that one case” that could lead to a new outbreak in an otherwise disease-free community, he says. “You need the vaccine to go the last mile.” 18. # Shrinking the Malaria Map From the Outside In 1. Leslie Roberts A debate is brewing over how best to fight malaria: Go for the quick wins, or the worst places first, or both? Richard Feachem wants to “shrink the malaria map.” By that, he and his Global Health Group at the University of California, San Francisco (UCSF), mean wiping out malaria at its “natural margins”—those countries on the edge of malaria transmission where the disease has just a tentative foothold—and working inward. It's going for the “low-hanging fruit,” he says. “It's a no-brainer.” That's not how some of Feachem's colleagues see it. No one questions the value of eliminating malaria, which means stopping disease transmission in a geographically defined area—where it's feasible today, although exactly what is feasible where and when, and what it might cost, is hotly debated. But many malaria hands worry that shrinking the map will divert money and attention from a far more urgent task: reducing the horrific toll of malaria in central Africa, where five countries account for 50% of all global deaths from the disease and elimination is not possible. In those countries, scaling up the use of existing tools—for instance, blanketing the population with insecticide-treated bed nets and ensuring fast, free access to malaria drugs—could have a huge payoff. Indeed, several countries, such as Zambia and Rwanda, have slashed cases by more than 50% (see p. 842). “There is not enough money for next year. So what investment would save the most lives and bring credibility to malaria control?” asks Carlos Campbell, who directs the malaria control program at PATH, a nonprofit in Seattle, Washington. “Do you work with Namibia and Botswana, which have very little malaria, and try to shrink the map, or do you use that money on bed nets for the DRC [Democratic Republic of the Congo]?” Feachem, who has assembled a large brain trust known as the Malaria Elimination Group, agrees that the bulk of donor funds and efforts should be concentrated in the highest burden countries. But he thinks the world can and should pursue both strategies. There is “absolutely no conflict” between the two, he says. He sees elimination as a piece of a three-part strategy that must be pursued simultaneously for “getting to a malaria-free world”: Control malaria in the “heartland,” eliminate it at the edges, and invest heavily in radical new tools that someday might make it possible to wipe the disease off the face of the earth. Many countries have already eliminated malaria—Canada and the United States, for example, and all of Western Europe. Several others are in the process, including Algeria, China, and Mexico, all on the northern boundary of transmission, and Argentina and the South Pacific island of Vanuatu on the southern boundary. In these places, the environment is less conducive to malaria transmission. At least as important, the countries tend to be high- or middle-income, with sufficient funds, political will, and strong enough health systems to devote to chasing the last few malaria cases and ensuring the disease doesn't come back. Feachem and the Malaria Elimination Group have published a 200-page prospectus, Shrinking the Malaria Map, that is essentially a how-to manual for getting local transmission to zero. In it they optimistically identify 39 countries as “eliminators” on the margins (see map, above). Some of them, such as Haiti and the Dominican Republic, will be at best an enormous challenge (see sidebar). Getting from low to zero transmission anywhere is a big leap, Feachem says. In addition to continuing high-level control with existing tools, it requires an exquisitely sensitive surveillance system to detect and treat every case and monitor for the parasite's return once it has been eliminated. Reintroduction is a huge threat—success totally depends on your neighbors, explains Feachem. That's why islands are considered the lowest of the low-hanging fruit. Four countries at the southern tip of Africa, where malaria transmission is now low, illustrate the challenges. South Africa, Swaziland, Botswana, and Namibia have declared their intention to eliminate malaria by 2015. The ambitious goal has been met with “skepticism,” says Oliver Sabot of the Clinton Health Access Initiative, but CHAI, in partnership with the UCSF Global Health Group, is helping those four countries achieve it. South Africa and Swaziland are making steady progress, Sabot says; Namibia and Botswana less so. Namibia boasts a strong health care system and good malaria control, he says, but it shares a border with Angola, which “has virtually no health care system and no control, and there is nothing to stop malaria from flowing across the border.” Robert Snow, who heads the Malaria Public Health and Epidemiology Group in Nairobi, is one of the skeptics. “Elimination costs a huge amount of money. Some of those are very poor countries. Is it an appropriate goal given the other health problems they face?” he asks. Sabot contends that shrinking the map gives countries that have already controlled malaria a much-needed goal: “The whole idea of shrinking the map is to give them a future to look to.” But even he worries that some countries, such as Ghana and Nigeria, are jumping on the elimination bandwagon prematurely, “without an awareness of what it actually means” either technically or financially. “From what we see so far, the expectation that a high-burden country can dramatically reduce funding for malaria once elimination is achieved is often false.” Everyone agrees that “before going down the chute” of elimination, each country should conduct a rigorous feasibility assessment, says Robert Newman, head of the Global Malaria Programme at the World Health Organization (WHO). Unfortunately, he says, the data needed are sorely lacking. He and other experts laud Feachem's group and CHAI for leading the way in collecting them, and they point to a recent analysis CHAI and others did with the government of Zanzibar as the gold standard. Zanzibar is one of the poster children for aggressive malaria control. With donor help, the islands massively scaled up interventions and reduced transmission dramatically. Now the government is grappling with what to do next, says Sabot: concentrate on sustaining those gains or try to stop transmission on the islands altogether? The CHAI analysis concluded that malaria elimination in Zanzibar is technically feasible but would be “very challenging” both operationally and financially. “It would likely be substantially more costly than maintaining control” for the next 20 years, Sabot says. What tipped the scales is the risk of reintroduction. Although it's surrounded by water, Zanzibar is very close to mainland Tanzania, where malaria transmission is robust, and “there's a ferry every hour,” says Newman. And once the parasite got in, it would quickly resurge, given the efficiency of the vector there. “Many of us were surprised the feasibility assessment said that it is not practical in the near term,” says Newman. “And if an island setting shows elimination is challenging now, it will be challenging in other areas.” Newman says WHO plans to work with the Malaria Elimination Group to develop evidence-based guidelines to help countries decide whether to move from control to elimination. “Data and assessment should drive public health policy. You don't want to set policy on enthusiasm, but you don't want to rain on it either.” 19. # Elimination Meets Reality in Hispaniola 1. Leslie Roberts* Hispaniola's fledging program to eliminate malaria, now sidetracked by the earthquake, illustrates the difficulties that arise when ambitious goals collide with reality. There is absolutely no reason for malaria to persist on the island of Hispaniola, says Donald Hopkins, longtime disease fighter and vice president for health programs at the Carter Center in Atlanta. All the other islands in the Caribbean have rid themselves of this mosquito-borne disease. And the Dominican Republic (D.R.), which shares the island with Haiti, has driven cases to remarkably low levels. By many counts, the country is well on the path to elimination, its stated aim. But it can't get there unless Haiti, where malaria is rife, achieves the same goal. Hispaniola's fledging program to eliminate malaria, now sidetracked by the earthquake, illustrates the difficulties that arise when ambitious goals collide with reality. In 2006, Haiti reported 33,000 confirmed cases of malaria; experts at the U.S. Centers for Disease Control and Prevention (CDC) peg the number at closer to 200,000 and warn that it may surge as the million or so people displaced by the 12 January earthquake remain in crowded camps with no protection from mosquitoes and the rainy season upon them. The D.R., by contrast, reports about 3000 cases a year, a number believed to be fairly accurate, and those are found mostly along the border with Haiti and among migrants from that impoverished country. Still, the D.R.'s success in slashing cases—and the successes of the neighboring islands in wiping out the disease—show that malaria elimination on Hispaniola is technically feasible. The biology is certainly obliging. Hispaniola is one of the few places in the world where the malaria parasite is still susceptible to chloroquine, a cheap and safe drug. Just one parasite, Plasmodium falciparum, transmits malaria on the island, as opposed to several in Africa, and although it causes the most lethal form of malaria, it is relatively susceptible to treatment. There's just one vector, too—Anopheles albimanus, or “white hands,” because of its distinctive white legs—and it is not very good at transmission, explains Angel Solís, an entomologist working with the Dominican health ministry. Unlike A. gambiae, the main vector in Africa, albimanus does not preferentially bite humans; a cow makes a fine meal. Nor does it prefer to bite indoors, where humans congregate—all factors that contribute to relatively low transmission there compared with Africa. Finally, as an island, Hispaniola is protected by vast expanses of ocean, with no other malaria-endemic countries nearby. So if the parasite can be eliminated, chances are low that it will get back in. Richard Feachem and colleagues at the University of California, San Francisco, have flagged Hispaniola as an “eliminator” on their malaria map (see main text, p. 849). And in 2006, the International Task Force for Disease Eradication (ITFDE), an expert group Hopkins chairs at the Carter Center, declared malaria elimination from Hispaniola “technically feasible, medically desirable, and … economically beneficial.” But there is more to elimination than favorable biology, says Robert Newman, head of the Global Malaria Programme at the World Health Organization. A country or region must have a strong health system, lots of money, and a government or donors willing to invest in surveillance long after the disease has disappeared. “All the stars must be aligned,” he says. Even before the earthquake, Haiti's malaria program, and even its public health system, essentially existed in name only. “There is a lack of nearly everything in Haiti,” says Nguyen-Dinh Phuc, a malariologist recently retired from CDC who has worked extensively on the island. “For malaria, they don't have enough people trained in microscopy; they don't have Ph.D. entomologists, or epidemiologists with expertise in malaria surveillance.” Even so, after ITFDE's proclamation, the Carter Center decided to see if it could prod the two countries to collaborate to eliminate malaria and another mosquito-borne disease, lymphatic filariasis (Science, 5 February, p. 634). The odds were against them, as elimination requires cross-border cooperation. “There was a lot of wariness between the two countries,” says Hopkins. “The impetus to collaborate was not strong.” But the benefits far exceeded the risk of failure, Hopkins says. With a small amount of money and a considerable amount of technical expertise, the Carter Center helped launch a pilot project in 2008 in the border towns of Dajabón, D.R., and Ouanaminthe, Haiti, which straddle the infamous Massacre River. The location, the site of one of the bloodiest encounters between the two countries, was chosen deliberately: Not only is it a hot spot of transmission, but experts reasoned if the two could collaborate there, they could anywhere. (Click here for a Carter Center video on the project.) As part of the pilot project, Dominican malaria experts have been giving their Haitian counterparts a crash course in the malaria-fighting techniques the D.R. has long employed. They include active surveillance to track down every single malaria case; the replacement of presumptive treatment, in which any person with fever is treated for malaria, with confirmed laboratory diagnosis; and fast and free treatment with chloroquine to address the illness and primaquine to help reduce disease transmission. The campaign also includes vector-control measures, such as bed nets and insecticide spraying, and monitoring for any signs of drug resistance. Within a year and a half, the project, led by David Joa Espinal on the Dominican side and Joanel Mondestin on the Haitian, had trained 10 community health workers, distributed 1200 bed nets, hired and trained three microscopists, and bought four microscopes, 11 motorbikes, two computers, and 37 cell phones. It was such a success, albeit a modest one, that in October 2009, with former President Jimmy Carter at their side, the health minister of each nation unveiled a binational plan to scale up these efforts across the entire island. The price tag:194 million over 10 years. Both presidents endorsed the plan, and Carter vowed to lobby the donor community for support.

Then the earthquake hit, and more urgent needs claimed priority. But as Haiti recovers, Hopkins and colleagues hope the idea of eliminating malaria and lymphatic filariasis can be revived.

The first order of business is to build a functioning public health system, with malaria control as a key component, says Larry Slutsker, who heads the malaria branch at CDC. Since the earthquake and several malaria outbreaks, CDC has had a team of malariologists on the ground to try to establish the rudiments of a control program in Port-au-Prince.

“Let's keep the elimination goal—it is great to have a plan, and benchmarks—but let's be realistic about what needs to be done first,” says Slutsker.

• * Leslie Roberts visited Dajabón and Ouanaminthe on a reporting trip in October 2009.