# News this Week

Science  21 May 1999:
Vol. 284, Issue 5418, pp. 1242
1. SPACE

# European Ministers Back Commerce Over Space Science

1. Helen Gavaghan*
1. Helen Gavaghan is a writer in Hebden Bridge, U.K.

BRUSSELS—The European Space Agency (ESA) veered sharply toward support for the commercial exploitation of space at a meeting of ministers from its 14 member states here last week. The abrupt shift left space science programs in the lurch. During the frequently tense meeting, ministers negotiated which elements to back from a $6.4 billion menu of programs proposed by the agency. They enthusiastically pledged funding for programs in satellite navigation, Earth observation, and communications. But, to the frustration of space scientists, the politicians voted to maintain a cap on science spending, which has already been in place for 4 years, and only reluctantly backed spending plans for the utilization of the space station. David Sainsbury, Britain's science minister, who chaired the meeting, declared it a success because of its strong support for commercial and environmental programs. Germany's research minister, Edelgard Bulmahn, told Science she was “very satisfied” with the outcome. ESA officials were not so buoyant. Commenting on a 10% downgrading of the Earth-observation budget, David Southwood, who headed the agency's efforts to develop the program, says the decision was “challenging” and “would concentrate the mind.” Others were less circumspect. “Hardly a day goes by when we are not asked for cost cuts,” says Jean-Jacques Dordain, head of group strategy and space development for the agency, “so this is business as usual.” View this table: The member states' changing priorities became obvious in the debate over the agency's science budget. Science is a “mandatory” program, which means that once a level of funding is agreed by unanimous vote, each country's contribution is determined by the relative size of its gross domestic product. Germany is the largest single contributor, paying 25%, then France at 17%. ESA's science directorate has often been praised for its coordinated approach to funding a steady stream of large and medium-sized missions put forward by the scientific community. But at the last ESA ministerial meeting, in Toulouse in 1995, ministers capped the science budget. “Because the program is mandatory,” says Hans Balsiger, head of ESA's advisory Science Policy Committee, “and there must be a unanimous vote, we are an easy target [for cuts].” The science directorate's difficulties were made more acute in 1996 when the inaugural launch of the agency's Ariane 5 rocket exploded, destroying Cluster, a mission to explore the Earth's magnetosphere with four spacecraft. At the Brussels meeting, the science directorate requested just over$2 billion for 1999–2003, which would bring the science budget's purchasing power back to 1998 levels by 2002. However, Karl Reuter, head of the ESA director-general's cabinet, told reporters that this turned out to be “one of the most difficult areas of discussion.” Germany, supported by France, wanted the budget freeze to continue. According to observers in the closed meeting, Bulmahn argued that the science directorate was amply funded and should have realized long ago that it could not do all that it wanted.

ESA had already warned the ministers that without any increase for inflation, the science directorate would not be able to maintain its lineup of currently approved missions, including the infrared and submillimeter telescope FIRST/Planck, due for launch in 2007, and the newly approved Mars Express, Europe's contribution to the highly coordinated stream of U.S. and Japanese missions heading for the Red Planet. In a compromise brokered by Sainsbury, ministers maintained the freeze on space science spending but approved a one-time payment of $44 million to safeguard Mars Express. “We should be grateful,” says Balsiger, “for the [$44 million], but I resent some of the statements made, because we had worked very hard [to cut costs].”

Asked what impact the compromise would have on Europe's space science policy, ESA Director-General Antonio Rodotà said immediately after the meeting: “There is no real answer that the agency can give today. The agency will have to turn to the Science Policy Committee. It is their problem to decide what is the best way forward.” The committee was due to meet this week, after Science went to press. “The questions we have to answer are, how many of the small missions do we have to cancel and how long can we defer FIRST/Planck without affecting the science or pushing up the cost,” Balsiger says. Some particularly vulnerable missions include Europe's part of the Next Generation Space Telescope and MiniStep, a fundamental physics mission to test whether inertial and gravitational mass are identical.

Although space science had to fight hard just to stand still, ESA's input to the deeply unpopular space station took an even more severe battering. The ministers could not reach agreement on funding for the crew return vehicle (CRV), a joint NASA/ESA craft that would ferry station crew to Earth, and so put off a decision on whether to join NASA in developing the vehicle. ESA had also proposed to spend about $107 million on microgravity research between 2000 and 2003 as a precursor to science aboard the station, but ministers slashed this program by half. Even ESA's basic program for space station exploitation only barely won full commitment of funds from member states. Germany, which has accounted for 41% of Europe's contribution, had long been the station's champion, but the new Social Democrat administration elected last year is no friend of the project (Science, 5 February, p. 784). Bulmahn says it is time to find ways to interest industry in exploiting the facility. “Industrial involvement on [the station] will be an important task for ESA,” she says. Speaking after the meeting, Rodotà told Science, “We will stick with the commitments we have made. That is the right decision, but we do not like this station. It is hard to say what the value of such a project is.” When the discussion turned from science to application programs, however, the ministers got out their checkbooks with relish. ESA put forward a new plan for a constellation of navigation satellites called GalileoSat that would allow receivers on the ground to fix their position with millimeter accuracy—a European competitor to the U.S. military's Global Positioning System. As an optional mission, each government can pledge however much it wants to spend, but even the 1-year definition phase of GalileoSat was 50% oversubscribed at the Brussels meeting, and without even being asked some member states, including the usually tightfisted United Kingdom, indicated their willingness to fund the development stage. The agency's overhaul of its Earth- observation strategy in the 2 years since Rodotà's appointment also won support. Previously, ESA's remote-sensing satellites had been huge technology demonstrators, such as ERS-1 and ERS-2, which did a lot to open up the field of radar remote sensing for civilian users. “These were magnificent beasts,” says Southwood, but the launch of ESA's Envisat this year “will mark the end of the great technology demonstrators for Earth observation.” Southwood and Roger Bonnet, the agency's science chief, have now drawn up a 5-year program of small application and research missions designed to provide data to a wide cross section of Earth-observation users. The plan won plaudits at the meeting, yet ministers cut ESA's request by 10%. As one U.K. official said at the meeting, “the days of the big prestige projects are over.” 2. ARCHAEOLOGY # New Date for the Dawn of Dream Time 1. Carl Zimmer* 1. Carl Zimmer is the author of At the Water's Edge. LAKE MUNGO, AUSTRALIA—Looming over this dry lake bed is a crescent of ancient sand dunes, carved by the wind into miniature canyons and mesas. Built up over tens of thousands of years, the dunes preserve traces of fires made by Aborigines to cook golden perch they once caught in the lake, spear points they used to hunt the kangaroos and other game that crowded Mungo's shores—and now, a team of researchers claims, the oldest human remains ever found in Australia. In the June issue of the Journal of Human Evolution, Alan Thorne of Australian National University (ANU) in Canberra and his colleagues put the age of a skeleton from Lake Mungo at 62,000 years. If correct—and some experts are withholding judgment until they see the paper—the date would mean that human beings had reached Australia tens of thousands of years earlier than some archaeologists thought. It also has implications for the history of modern humans. Many researchers maintain that all modern humans descend from a single population of Africans dating back perhaps 100,000 years; these founders later spread out across the Old World, replacing any humans or hominids they encountered. One of the last places they would have reached would be Australia, so an early date for their arrival would mean an earlier migration than some researchers had pictured—or perhaps an alternative scenario of modern human origins. “If the dates are reliable, the implications are very substantial,” says Stanford University's Richard Klein. The skeleton itself, called Mungo 3, is not a new discovery. In 1974 geomorphologist James Bowler of ANU uncovered the remains of a lightly built man in one of the dunes. Thorne completed the excavation and determined that the body had been ceremonially buried in a grave, with red ocher scattered over it and its fingers intertwined around its penis. Carbon-14 dating initially put its age at 28,000 to 32,000 years old. But in the mid-1980s, as researchers further refined 14C dating techniques, they were able to remove younger organic contamination and push the date back to at least 38,000 years ago. Some researchers wondered if Mungo 3 might be still older. When organic material gets to about 40,000 years old, so much of its 14C has decayed that it's often difficult to get a precise age. And new dating techniques that can reach beyond 14C had already hinted that human beings were on the continent well before 40,000 years ago. These methods rely on counting the electrons knocked into defects in a mineral's crystal structure by natural radiation. Called thermoluminescence and optically stimulated luminescence (OSL), they can measure how much time has passed since a material has been heated in a fire or exposed to sunlight, which “zeroes” the clock by allowing the electrons to fall back into place. Starting in 1990, Richard Roberts of La Trobe University in Melbourne and his colleagues used these techniques to date several rock shelters in northern Australia to between 50,000 and 60,000 years old. But some archaeologists remained skeptical about these old dates, because the techniques are tricky to use and have sometimes produced outlandish ages (Science, 10 October 1997, p. 220). So Thorne and his colleagues decided to date Mungo 3 not only with OSL but also with two other unrelated techniques. Geochronologist Reiner Grün, also from ANU, used a method called electron spin resonance (ESR), which counts electrons trapped in defects in the crystal structure of bone or tooth. Grün also measured how much of the uranium in the Mungo 3 skeleton had decayed into two daughter elements, thorium and protactinium. Normally researchers have to destroy a sample of bone to get these measurements, but Grün put the Mungo 3 cranium into a custom-built lead-lined box; for over a month he then counted the gamma rays that flew from the bone as the uranium decayed. Meanwhile, ANU physicist Nigel Spooner took samples of the sand from around the skeleton for OSL measurements. The three methods gave pretty much the same result: 61,000 ± 2000 years from OSL, and 62,000 ± 6000 years from both the uranium-series and ESR methods. This new estimate of the age of Mungo 3 is 50% older than any previous date for Australian human remains. Thorne and his colleagues are confident in the date. “Here is a case where three different methods provide extremely similar results,” says Thorne. Some experts remain cautious, waiting to see the full details. Roberts notes that “uranium is a very mobile element in groundwater and can enter and leave a deposit at will,” skewing its apparent age. And he adds that “ESR suffers from some of the same problems, so it's not a truly independent age comparison.” But if the date does hold up, it may challenge some versions of the out-of-Africa theory. Researchers such as Klein have argued that the modern humans emerging from Africa brought with them art, ritual burials, and other signs of cognitive sophistication. The oldest widely accepted evidence for this “human revolution” in Africa is 50,000 years old, and 40,000 years in Europe (Science, 20 November 1998, p. 1451). But the ceremonial burial of the Mungo 3 man, along with the boat-building skills that got his ancestors to Australia in the first place, point to such sophistication at a much earlier date. If the date is real, says Klein, “it would mean either that there was a separate evolution of modern humans and modern human behavior in the Far East, or that modern humans emerging from Africa somehow managed to reach the Far East at least 20,000 years before they reached the Far West.” Thorne prefers the first alternative. He points out that whereas Mungo 3 had a slender, gracile build, more recent skeletons, dating back only 15,000 years, were of a very robust build. Thorne suggests that a gracile population evolved in east Asia and came to Australia before 60,000 years ago, while a more robust one came later from southeast Asia. The two peoples then interbred to create today's Aborigines. But Chris Stringer, a paleoanthropologist at the Natural History Museum in London, doesn't see any need to give up on the out-of-Africa picture, even if the first Australians didn't look like their successors. To him the robust anatomy in later Australians could have been the result of local evolution on the continent after Mungo 3. “There is no evolutionary reason why populations cannot become larger and more robust through time,” says Stringer. Either way, Mungo 3 may be a crucial clue to the fate of the 2-ton wombats, 3-meter kangaroos, and other giant animals that once inhabited Australia. Last January Gifford Miller of the University of Colorado, Boulder, and his colleagues offered some circumstantial evidence that humans were to blame for their disappearance. Studying eggshells, they found that a giant flightless bird called Genyornis abruptly went extinct about 50,000 years ago. Climate records show no drastic change at the time, leading Miller to suggest that the culprit was hunting or an ecological collapse triggered by humans. Aborigines regularly set fires for everything from flushing out game to clearing water holes. In the process, they may have destroyed fire-sensitive plants and driven the animals that depended on them to extinction (Science, 8 January, p. 205). “If humans didn't come till 40,000 years ago, they aren't involved and we have to give up on this hypothesis,” says Miller. “That's why this date on Mungo is important. It's really showing people are not only present in Australia but in the interior. We've got people basically everywhere 60,000 years ago.” Thorne isn't persuaded that the extinctions mark the arrival of humans, because he thinks people may have been in Australia long before the Mungo 3 man lived. “We do not know when humans first arrived, and this date on Mungo 3 is an important way of saying that.” 3. CLINICAL RESEARCH # Shutdown of Research at Duke Sends a Message 1. Eliot Marshall Last week was a roller coaster for Joanne Kurtzberg, a transplant researcher at Duke University's Medical Center in Durham, North Carolina. First she heard—from a nurse—that her study using placental cord blood as a stem cell source had been suspended. “One of the OB-GYN doctors had come down the hall and said we couldn't consent any more patients,” Kurtzberg recalls. Most other clinical researchers at Duke received similar shocks as word raced around the building that the Office for Protection from Research Risks (OPRR)—a federal watchdog agency that is part of the National Institutes of Health in Bethesda, Maryland—had lifted the center's authority to do federally funded research. Four days later, the immediate crisis eased when OPRR lifted the sanctions. But the reverberations will continue at Duke and elsewhere. OPRR had suspended the Duke studies after months of urging administrators to correct “serious deficiencies” in procedures for monitoring consent and keeping records. Although the agency has now accepted a reform plan put together by chancellor for health affairs Ralph Snyderman and medical school dean Edward Holmes, OPRR director Gary Ellis notes that Duke still needs to re-review many projects for compliance with human subjects protection rules. That could take weeks. And with this action—the second shutdown of research at a major clinical center it has ordered in as many months—OPRR has put every federally funded U.S. research institution on notice that its right to conduct clinical research could be summarily yanked. As though to underline the warning, the President's National Bioethics Advisory Commission (NBAC) also released a statement last week identifying defects in the U.S. system of protecting human research subjects. In a 4 May letter to President Clinton, NBAC chair Harold Shapiro describes gaps in enforcement and promises to deliver a “comprehensive report” in a few months loaded with recommendations for improvement. OPRR isn't waiting for such advice. In focusing on Duke, it has targeted one of the nation's top biomedical centers. Duke has been aggressively adding staff and facilities —including a special center for industry-funded studies called the Duke Clinical Research Institute—to increase the number of clinical trials it manages. It now has 2000 in its portfolio, says medical center spokesperson Nancy Jensen. In December, OPRR sent nine experts to take a close look at Duke's procedures for monitoring the welfare of patients participating in this burgeoning research enterprise. After poring over records and interviewing staff, the team found no evidence that any patients had been harmed but concluded that the institution needed to tighten its practices. Immediately after the site visit, OPRR sent a six-page letter to Snyderman listing 22 problems and asking for major changes. In February, OPRR followed up with another letter, warning that fixes offered by Duke were “unsatisfactory.” For example, OPRR had asked that members of a panel that clears proposed experiments—the Institutional Review Board (IRB)—be better educated on relevant federal law. Duke responded by suggesting that it might invite IRB members to an annual lecture and brief them on useful Internet sites. Not good enough, OPRR said. Instead, it called for “provisions to ensure that all IRB members periodically receive interactive or didactic training from expert consultants working in the field of human subject protection.” OPRR also faulted Duke for allowing administrators to serve in a conflicted role as voting members of the IRB, for keeping sketchy records, and for failing to staff the IRB adequately. Duke responded in April and May with more suggestions, but OPRR officials again failed them. On 10 May, citing a “lack of progress” and a “failure of leadership,” OPRR halted clinical trials. Snyderman, who says he had no inkling OPRR was considering such drastic action, estimates that “hundreds” of trials were affected. Some clinicians were faced with the possibility that they would have to delay enrolling patients while their projects were re-reviewed. Kurtzberg, for example, learned that her study offering experimental therapy to infants with a neurological disorder may fall in that category. She's still not sure when enrollment will resume. Duke's medical chiefs may take comfort in knowing that they are not alone. Last month, OPRR shut down the Veterans Administration hospital in Los Angeles for similar reasons, and 5 months earlier, the Rush Presbyterian St. Luke's Medical Center in Chicago (Science, 2 April, p. 18, and 6 November 1998, p. 1035). Asked if this signals an escalation in enforcement, perhaps in response to congressional urging, OPRR chief Ellis responded, “that's for others to comment on.” However, he noted that at a hearing last year, a prominent congressman called OPRR's enforcement efforts “pathetic” and “absurd.” Ellis added, “With intense interest in human studies from Congress, the President's National Bioethics Advisory Commission, and advocacy groups, this is no time for dawdling.” 4. PLANETARY SCIENCE # Space Rock Hints at Early Asteroid Furnace 1. Erik Stokstad Long before Earth or any other planet had formed around the sun, a vast cloud of dust began to coalesce into asteroids. Most of the drifting chunks were the kind of stone-cold rubble that Han Solo had to weave through in his clunky old spaceship in Star Wars, but others were big enough, and hot enough, to ooze lava. For decades, planetary scientists had suspected that radioactive decay stoked the furnace of these hefty asteroids, some of which later merged into planets. But the embers that would identify the heat source have long since died out. Now comes the first hard evidence of what melted large asteroids in the early solar system. On page 1348, a team led by planetary scientist Gopalan Srinivasan of the Physical Research Laboratory in Ahmedabad, India, reports that a 4.57-billion-year-old meteorite—a fragment of an asteroid that developed a molten interior and crust in the early days of the solar system—bears the unmistakable signs of a radioactive heat source. The rock once contained enough of the radioactive isotope aluminum-26 to have melted. “They've got the smoking gun,” says Stuart Weidenschilling of the Planetary Science Institute in Tucson, Arizona. Thought to have been blown into our solar system on the winds of a nearby supernova explosion or, perhaps, spawned by particles from the young sun, 26Al was first proposed as a heat source for the early solar system back in 1955. With a half-life of 730,000 years, 26Al could have melted early asteroids, then disappeared long before our own planet grew to full size. But it wasn't until the mid-1970s that researchers found indirect evidence for 26Al's existence: the presence of its decay product, magnesium-26, in calcium-aluminum-rich inclusions, the first specks thought to have formed in our solar system's primordial gas cloud and preserved in ancient meteorites. But researchers came up empty-handed when they looked for 26Mg in meteorites from parent asteroids that once had molten interiors. Complicating the search, these so-called differentiated meteorites make up fewer than 5% of those that hit Earth. Lucky for Srinivasan and his colleagues, just such a meteorite thundered into the desert state of Rajasthan in western India on 20 June 1996. Called Piplia Kalan after a nearby village, the 42-kilogram meteorite resembles basalt, and its tiny crystals suggest it cooled rapidly after melting. Auspiciously, one section contained crystals of plagioclase, an aluminum-laden mineral that might once have been rich in 26Al. And compared to most other differentiated meteorites, the grains contained little magnesium. That led Srinivasan and his team to think they had a good shot at finding the 26Mg produced by 26Al decay, which would be swamped by common magnesium in most plagioclases. Indeed, 26Mg levels in four grains of Piplia Kalan were up to 3% higher than the usual amount in terrestrial plagioclase. By cosmic chemistry standards, says Srinivasan, “this excess is very significant.” The finding “strengthens implications that 26Al was the heat source” at the heart of asteroids, says Glenn MacPherson, a geochemist at the Smithsonian Institution in Washington, D.C. For connoisseurs of asteroid history, it also suggests how long it took for the parent body of Piplia Kalan to melt and cool after the formation of the solar system. Like measuring time according to sand in an hourglass, our solar system's initial allotment of 26Al can be extrapolated from the 26Mg in calcium-aluminum inclusions. The tricky part is that in molten rock, the hourglass wouldn't collect falling sand, so to speak, because any 26Mg would have been elbowed out of minerals that prefer aluminum atoms. After the molten rock cooled into crystals, however, the 26Mg would become trapped and begin to accumulate; its abundance would reveal that of 26Al during crystallization. By comparing 26Mg abundance in the calcium-aluminum inclusions to the vastly smaller amount in Piplia, the team estimates that 5 million years must have elapsed before the plagioclase in Piplia crystallized. This time span for accreting and melting the parent body jibes with computer models of the process, providing “a real shot in the arm for theoretical work,” says geochemist Richard Carlson of the Carnegie Institution in Washington, D.C. The hunt is on for longer lived isotopes, such as samarium-146, manganese-53, and iron-60, that may have been trivial heat sources in the early solar system but, by their abundance in differentiated meteorites, could help narrow the window on when asteroids began solidifying. Such radiometric dates “will help explain processes operating 4.6 billion years ago in the inner solar system,” says Srinivasan. And that, notes Carlson, could help us better understand modern features of our solar system, such as the chemical composition of different planets. 5. ENVIRONMENTAL POLICY # EPA's Piecemeal Risk Strategy on Way Out? 1. Jocelyn Kaiser In diagnosing the ailments afflicting the Florida Everglades, researchers at first painted phosphorus as the archvillain: The nutrient, they concluded, nurtured the cattails that choked the saw grass and sent many species into decline. Among the staunchest advocates of this message were Environmental Protection Agency (EPA) scientists, who since the 1970s had trained a harsh light on phosphorus in U.S. estuaries. But when experts from several dozen agencies began in the last decade to plot a strategy for restoring the Everglades' natural plumbing, they discovered that phosphorus's impact on the ecosystem is dwarfed by the effect of dredging canals and other large-scale physical disruptions of water flow. The EPA's tunnel vision in the Everglades is just one example of how the agency sometimes fails to look at the bigger picture when assessing risk, says Mark Harwell, an ecologist at the University of Miami. But that may soon change. Earlier this month, a blue- ribbon panel released a draft report calling on EPA to broaden its outlook by assessing whole suites of chemicals and other threats to health and ecosystems, not just single pollutants. “This really calls for a big change,” says Joan Daisey, a physical chemist at Lawrence Berkeley National Laboratory in California and one of 49 members of an EPA Science Advisory Board (SAB) panel that spent 3 years on the report. In a 1990 report, the SAB urged the agency to set priorities by ranking risks according to scientific reviews rather than mandates from Congress and lawsuits. Laws often focus on single pollutants in specific media, such as air or water, and EPA has tended to hew to the boundaries set by the legislation, despite having leeway to take a broader approach to analyzing risks and crafting policy. Many scientists argue that EPA's ability to improve environmental quality through this pollutant-by-pollutant tack is diminishing, much like a curve nearing an asymptote, says Harwell, a report author. To help steer the agency toward a more wide-ranging approach to environmental threats while carrying out mandated legislation, EPA Administrator Carol Browner and Congress in 1995 asked SAB to update its risk report. The panel's latest offering, called Integrated Environmental Decision-Making in the 21st Century, has a two-part remedy for EPA. Half the prescription is to probe the breadth of risks to human health—suspected carcinogens and hormone mimics in drinking water, say—or to an ecosystem. The other half is to investigate a “broader range of risk reduction options,” which boils down to providing a larger menu for regulators: from strict caps on industrial emissions to guidance on how the public can voluntarily avoid certain risks. The report lauds a few good models, such as a 1996 law that instructs EPA to assess the health risks of pesticide combinations. New efforts, the report states, could include examining the range of pollutants of particular risk to city dwellers, or helping an industry devise a better strategy for reducing overall risk from its emissions. The report offers EPA a load of suggestions on the finer points of this new direction: The agency should invest more in gathering data on the universe of potential risks and in the social science expertise needed to figure out how best to address them, for example, and work more closely with other agencies on issues that tend to fall between the cracks, such as the threats to ecosystems from tampering with water flows. EPA officials declined to comment on the report, which still must undergo peer review. The SAB's recommendations are a tall order for an agency still groping for a system with which it would routinely use the best science in setting policy. And the report is no panacea. “It's not as big a step as I think it could have taken,” says economist Paul Portney of Resources for the Future, who wanted to see some examples of how EPA might weigh a range of health threats—for example, whether the cheapest way to cut overall risks from air pollution might be to tackle nitrogen oxides rather than particulate matter. But most agree that the report will help to set EPA on a path toward heeding science more regularly. “This is the kind of thing the scientific community has been calling for,” says Bailus Walker, a toxicologist at Howard University in Washington, D.C. “It's good to see the SAB is giving it a nudge.” 6. AUSTRALIA # Budget Backs Report on Boosting Biotech 1. Elizabeth Finkel* 1. Elizabeth Finkel writes from Melbourne. MELBOURNE—With 4 years as research director of California Biotechnology under his belt, Australian molecular biologist John Shine returned home in the late 1980s hoping to apply his knowledge of the burgeoning biotech industry to a start-up company based on the cloning of a key brain neuropeptide receptor. But the company, Pacific Biotechnology, couldn't survive in Australia's notoriously conservative investment climate: Within 3 years it had folded from a shortage of capital, one of several essential steps in a process of turning academic breakthroughs into products. Last week the Australian government announced a new budget that scientists say signals its desire to change that climate. Its plan to more than double spending on basic medical research over 6 years—to$235 million by 2004–05—mirrors a recommendation last December by a high-level review panel looking at ways to improve the country's competitiveness in health and medical research. In addition to the government's speedy adoption of one of the report's three recommendations, other parts of the budget address its call to strengthen health management. Scientists are optimistic that support for the third leg of the stool—reform of tax laws to encourage venture capitalism—is not far behind. And they give much of the credit to the review panel's chair, businessman Peter Wills, and to Health Minister Michael Wooldridge.

“They are the two big W's,” says Shine, who is now director of Sydney's prestigious Garvan Institute. “Wills and his committee did an outstanding job of painting a picture of the economic value [of investing in biotechnology], and we had a committed [health] minister. If either one hadn't been there, we would have lost the opportunity.”

The review, which Wooldridge commissioned, described the need for a comprehensive strategy to create a robust health-care and biotech industry. It called for boosting the size and duration of grants from the National Health and Medical Research Council (NHMRC), as well as raising salaries and allowing researchers greater mobility. It said there was no mechanism to set priorities in public health research, and it urged a reduction in the capital gains tax, at 48% one of the world's highest. The resulting lower return stifles investment in companies like Shine's, leaving him to resurrect his research at the Garvan Institute and to seek help from global giant Bristol-Myers Squibb in preparing the drug for market. “If we'd had the high-risk venture capital, we'd be looking at a 10% royalty on the final drug instead of 1%,” he says.

Although the government has not formally responded to Wills's review, it is clear that officials have embraced many of its tenets. Apart from the bigger budget, Treasury Minister Peter Costello announced plans to transform the NHMRC from a voluntary group with a tiny staff to an organization with a full-time leader who also advises the government. A global search will be held for a high-profile chief executive. In addition, the government named Wills to lead a new committee to implement his recommendations. Elsewhere in the budget, the government commits $13.4 million to operate a new genome research facility that will allow Australian researchers to participate in the international mouse sequencing project and other major efforts, as well as to identify disease genes in Australian pedigrees. Such government largess does not extend to all fields, however. Indeed, nonmedical university-based research is slated to drop 5.5% in the next 2 years as part of a 2.3% drop in overall spending in higher education. “This policy is undermining the government's entire innovation strategy,” says Australian Academy of Science president Brian Anderson. Pointing to an allocation of$11.7 million for a national biotechnology strategy to cover crops, animals, and humans, Anderson argues that, “for that to produce practical outcomes, you've got to put universities in a stronger position.” Advocates of university research note bitterly that the government has not yet responded to a high-level report on higher education released more than a year ago, while Wills's review has apparently been embraced barely 5 months after its submission. Even some of the increases ring hollow: A promised $60 million for university infrastructure, for example, merely brings things back to 1996 levels, Anderson notes. Despite those concerns, researchers across fields applaud the windfall for medical research as an example of what Australia must do to thrive in today's knowledge economy. Says Wills, “This [is] a historic commitment. … Now we will be working to get that all in place and start the process.” 7. ECOLOGY # New, Nonchemical Pest Control Proposed 1. Anne Simon Moffat In a springtime ritual as old as the suburbs, millions of gardeners are spraying 2,4-D and other herbicides to cultivate perfect lawns, free of dandelions and other weeds. Rarely do those gardeners realize that they are usually applying more herbicide per square meter than farmers treating their fields. Now, research by a team led by ecologist David Tilman of the University of Minnesota, St. Paul, points to a new, more environmentally benign strategy for controlling pesky weeds that may help suburban lawn growers kick their chemical herbicide habit. In the spring issue of the quarterly Ecological Applications, Tilman, with Elizabeth Tilman (his daughter), also at Minnesota, and two researchers in the United Kingdom—Michael Crawley of Imperial College, Ascot, and A. E. Johnston of the Rothamsted Experiment Station in Hertfordshire—report that dandelions have an Achilles' heel: a high need for the mineral potassium. As a result, a lawn planted with grasses that don't need much potassium, such as bent grass, foxtail, and fescue, can be kept lush and green while dandelions remain in check, as long as potassium fertilizer isn't added. “I like the idea; it's novel, it should be pursued,” says Cornell University agricultural scientist David Pimentel. This strategy might also work to manage weeds on farms, he adds. The Tilmans first suspected that low potassium might limit dandelion growth in 1992. While visiting the Rothamsted Experiment Station, they noticed many dandelions growing on experimental plots that had received high potassium applications, while adjacent plots without such fertilization had few of the weeds, even though they were naturally seeded by the neighboring, prolific dandelions. This suggested that dandelion abundance depends on potassium fertilization and, presumably, on the outcome of competition for potassium with other species. Analysis of Rothamsted data further supported this idea, which the researchers subsequently tested in the greenhouse and, in 1996, on lawns in north Minnesota. Both studies confirmed their theory. In the greenhouse, the team planted dandelions and five kinds of grasses, then analyzed the tissues of the plants. They found that dandelions had by far the highest potassium content, suggesting that these weeds have a hearty appetite for potassium. They also found that in plots given a low-potassium fertilizer, the biomass of dandelions, fescues, and orchard grass dropped, but the dandelions had the greatest reductions—down to 81% of their biomass in plots given the complete mineral treatment. In addition, when the group studied 19 lawns that had not been fertilized, treated with herbicide, or hand-weeded for 7 years—a tough set of criteria to fulfill in suburban America—they found that the density of dandelions per square meter correlated with the tissue potassium content of the weeds, a further indication that dandelions need ample potassium to thrive. Based on these results, David Tilman suggests one step gardeners could take right now. Many common lawn fertilizers contain a healthy dose of potassium, encouraging the growth of dandelions and, subsequently, the use of 2,4-D and other chemical herbicides to kill them. For many lawns, a fertilizer of ammonium sulfate or ammonium phosphate only would be better, he says. One exception, however, would be for lawns of Kentucky bluegrass, which is almost as greedy for potassium as dandelions are. Still unclear is how widely applicable the strategy of controlling weeds through nutrient limitations will be. Gardeners don't just worry about dandelions, after all. They also have to deal with other weeds, such as crabgrass, and it remains to be seen whether potassium limitation will help control these lawn invaders. And although Pimentel suggests that manipulating competition for nutrients might also help control weeds on farms, he says it may work best when farmers want to favor just a few plant species. Fields in which different crops, usually with complementary nutritional needs, are rotated and pastures with 10, 20, or more different plants will be tougher to manage. Still, field researchers believe the current work is worth following up. Says Mississippi State University weed scientist David Shaw, “We're only beginning to understand the relationship between soil characteristics and weeds.” Once agricultural researchers learn more, dandelions may not be the only weeds subject to this subtle kind of control. 8. ASTRONOMY # Giant New Telescope Bags Gamma Ray Burst 1. Govert Schilling* 1. Govert Schilling is an astronomy writer in Utrecht, the Netherlands. Early last week, another of the remote, powerful explosions called gamma ray bursts flared in the southern sky, in full view of the Southern Hemisphere's largest telescope, the new Very Large Telescope (VLT) in Chile. As it faded, the burst's visible afterglow provided what may be the strongest support yet for a budding theory that these mysterious blasts emit radiation in two opposing beams, which makes them visible at great distances when one of the beams happens to be directed at Earth. The first glimpse of the burst, now named GRB990510, came from two satellites, NASA's Compton Gamma Ray Observatory and the Italian-Dutch satellite BeppoSAX, which caught a bright flash of gamma rays on Monday, 10 May, about 08.49 universal time. Simultaneous x-ray observations by BeppoSAX pinpointed the sky position of the burst near the celestial south pole, and within 10 hours, astronomers had spotted a relatively bright but fading optical counterpart with the 1-meter Sutherland telescope at the South African Astronomical Observatory. In the past, astronomers have relied on the 10-meter Keck telescope on Mauna Kea, Hawaii, to analyze such afterglows for clues to the distance of the burst. This one was located too far south to be seen from Hawaii. But astronomers can now get a comparable view of bursts in the southern sky with the VLT, commissioned just this spring. Titus Galama and Paul Vreeswijk of the University of Amsterdam, who are coordinating the follow-up observations of the burst, used Antu, the first of the VLT's four 8.2-meter telescopes, to collect a spectrum of the afterglow. The spectral lines had a redshift—a displacement caused by the expansion of the universe—of 1.61, implying that the burst took place more than halfway across the universe. Theorists believe that gamma ray bursts signal stellar cataclysms—the collapse of a giant star or the collision of a pair of neutron stars. Even so, their brightness at great distances has been a puzzle. The latest burst provides a clue, say astronomer Shrinivas Kulkarni and his colleagues at the California Institute of Technology in Pasadena. A day and a half after the burst, data from a 1.25-meter telescope at Mount Stromlo Observatory in New South Wales, Australia, showed that the visible light from the afterglow started to decrease more rapidly than before—a break that happened simultaneously at different wavelengths, Kulkarni says. “That's what you expect when a jet [of particles] moving toward you with almost the speed of light is slowing down,” Kulkarni says. Such a jet would channel radiation straight down its axis, producing a searchlight beam visible at enormous distances. Although another burst earlier this year had shown similar behavior (Science, 26 March, p. 2003), he says, “this burst shows even better evidence for beaming.” 9. SOLID-STATE PHYSICS # Picking Up Bits of the Electron's Charge 1. Alexander Hellemans* 1. Alexander Hellemans is a writer in Naples, Italy. Although every science student is taught that the fundamental, indivisible unit of charge is that of the electron, physicists are now busying themselves collecting even smaller bits of charge. Two years ago, teams in France and Israel both found that in layers of electrons exposed to high magnetic fields at low temperatures, charge could shatter into “quasi-particles” that had one-third the fundamental charge. In this week's issue of Nature, the Israeli team announces it has now spotted one-fifth-charge quasi-particles. The new results provide further confirmation for the theory put forward in 1983 by Robert Laughlin of Stanford University to explain the fractional quantum Hall effect, a phenomenon in which the tiny units of the quantum world have large-scale effects. Although the theory is widely accepted and Laughlin shared the 1998 Nobel Prize in physics for it, some physicists were not entirely comfortable until a fractional charge was positively identified. Says Laughlin, “I'm very happy” about the new results, “but I am not surprised.” The Hall effect, known since 1879, describes how applying a magnetic field perpendicular to a current-carrying wire creates a voltage across the wire's width, because the field causes the electrons to bunch up on one side. In the 1980s, physicists discovered that when electrons are trapped in a thin layer between two semiconductors at low temperature and high magnetic fields, the Hall voltage across the conductor increases in discrete steps rather than continuously. The size of the steps reflects the discrete charge of the electron or integer multiples of it. But to their surprise, physicists also discovered steps that could only be explained by fractions—or multiples of fractions—of that charge, 1/3, 2/3, 2/5, and 3/7. To explain those fractional charges, Laughlin proposed that quasi-particles form in the electron layer when the electrons team up with magnetic vortices, tiny whirlpools of magnetism. In very simple terms, the vortices bound to an electron repel other nearby electrons and in effect “shield” part of their electron's charge. This has the effect of making the quasi-particle appear to have only a fraction of the electron's charge. Quasi-particles made up of an electron bound to two vortices would give rise to fractional charges of one-third; and at slightly lower magnetic fields, even smaller fractional charges could appear. In September 1997, two groups, one led by Mordehai Heiblum at the Weizmann Institute of Science in Rehovot, Israel, and the other led by D. Christian Glattli of France's Atomic Energy Commission in Saclay, announced that they had found direct proof of the existence of one-third fractional charges in a quantum Hall setup (Science, 19 September 1997, p. 1766). Now the same team at the Weizmann Institute has refined its technique and spotted the more elusive one-fifth-charge quasi-particles. To sieve these quasi-particles from the electron layer, they directed the charges toward a very narrow channel. The channel allowed the one-third-charge quasi-particles through, while reflecting a small number of one-fifth-charge particles back. The researchers then detected the reflected quasi-particles and their charges by the impulses they create in a very low noise amplifier. Because the signal generated by the one-fifth-charge quasi-particles is much weaker than that of the one-third charges, the team had to improve its amplifier. “We had to build amplifiers that work at the quantum limit,” says Heiblum. Laughlin is convinced that with further progress in measuring technology, even smaller charge fractions may soon be discovered. “All of these other fractions are just born out of the first one,” he says. 10. SCIENTIFIC COMMUNITY # 'New Physics' Finds a Haven at the Patent Office 1. David Voss Dozens of recent patents have been awarded for devices that invoke principles outside accepted science, such as exotic nuclear physics and psychic forces A famous cartoon shows a man waiting outside the Patent Office with a complicated gadget in his lap. He looks over and sees another man holding exactly the same contraption. The image reflects a common myth—that the government checks that an invention relies on accepted principles before granting a patent. But consider two recent patents: 5,616,219 and 5,628,886, issued to Clean Energy Technologies Inc. of Sarasota, Florida, for an electrochemical device that is claimed to put out more energy than is possible by chemistry alone. Or take Clean Energy's patent 5,672,259, for a process to transmute radioactive elements by electrochemistry. Physicists who have examined these patents say the claims resemble cold fusion; the company rejects that label but says its products do exploit “new nuclear physics.” Either way, the devices would challenge some basic notions of modern physics if they worked as advertised. A cursory search of recent patents reveals dozens of others like them. Such patents confer prestige and legitimacy, attracting investment dollars and customers. “Having a patented device, with lots of fancy equations in the manual, that's partly why people take it seriously,” says Keith Conover, a University of Pittsburgh physician who has studied a patented instrument said to be able to find buried disaster victims by “dielectrokinesis.” Patents like those awarded to Clean Energies have also helped keep the cold-fusion field alive. Roundly scorned by the scientific establishment in the 10 years since Stanley Pons and Martin Fleischmann first said they had achieved fusion in a jar, claims of unlimited energy live on—albeit under other names—in the patent literature. Most disturbing to some onlookers is the window such patents offer on the patent examination process. The U.S. Patent and Trademark Office (USPTO) is now staggering under an onslaught of patent applications. Its nearly 3000 examiners must process roughly 240,000 intellectual property claims every year, a number that is increasing by more than 8% annually because of increases in software and biotech applications. Says one former USPTO employee, “They are desperate and they're hiring like crazy.” The office plans to add another 700 examiners in the coming year. And as Richard Maulsby, a PTO spokesperson, admits, “It is very difficult for us to do all this hiring and to maintain quality.” The result, in some cases, is inspectors who have little experience—or are themselves devotees of fringe technology (see sidebar on p. 1254). The Patent Office has long been besieged by inventors seeking patents on weird gadgets, and patent law is specifically written to restrict patents on one kind of device with perennial appeal, perpetual motion machines. Anyone who wants to patent such a mechanism has to submit a working model as part of his application. For most inventions, however, the bar is lower. Except for perpetual motion machines, “the Patent Office hasn't required a working model since the 19th century,” says patent attorney Michael J. Colitz, the creator of the Wacky Patent of the Month Web page. Instead, patent law requires only that an invention be novel, nonobvious, and reducible to practice. “Reducible to practice” sounds like “really works,” but by clever wording, patent applicants can dodge tough scrutiny of how realistic an invention is. The trick is to avoid the perpetual motion label and others, such as “cold fusion,” that might raise red flags for the patent examiner who searches prior patents and judges whether a patent claim makes sense. “Each patent examiner has different criteria,” says Colitz. And these days, sincere but poorly trained examiners are making many of those judgments, says patent consultant Greg Aharonian, editor of Internet Patent News. One examiner says he was interviewed over the phone by a supervisor. A few days later he got a package in the mail. “I thought it was an application,” says the interviewee. “But it was a form confirming my acceptance of the position.” The low salaries at the Patent Office don't help matters, say patent examiners and outsiders. “They have a variety of problems in not being able to retain good patent examiners because of the high salaries outside,” says Aharonian. “You really have to be a patriot to want to work at the Patent Office.” ## Cold fusion reheated? Inexperienced patent examiners may be one reason why some unlikely inventions—helped along by clever patent attorneys—have recently won patents. Although the Patent Office initially rejected cold-fusion patents after Pons and Fleischmann's memorable Salt Lake City press conference in 1989, some experts say the Clean Energy patents show that such patents are now slipping into the books. James Reding, Clean Energy's chief executive officer (CEO), insists that his company's technology is not “cold fusion,” although he says it does exploit nuclear processes. But every physicist Science has asked about the Clean Energy patents, including IBM's Richard Garwin and William Happer of Princeton University, says they describe what are essentially cold-fusion devices. And the March/April 1999 issue of Infinite Energy magazine, a publication for cold-fusion buffs, includes Clean Energy work in its list of “Key Experiments that Substantiate Cold Fusion Phenomena.” The patents say that the devices generate excess heat by passing a current through a cell containing beads coated with a metal such as palladium and exposed to various hydrogen isotopes—the same setting where cold fusion was said to occur. Garwin and others say the devices are unlikely to prove viable, either as energy sources or as systems for rendering radioactive waste harmless. Conditions in an electrochemical cell fall far short of what is needed to trigger nuclear reactions, they note. “The cell has never produced any excess heat, in my judgment,” says Garwin, who has looked at Clean Energy's data. “And this remediation of radioactive materials is incredible and has not been demonstrated.” Reding responds that he knows the physics is controversial, but “the technology is very real.” Reding says that the company's first attempts to patent the devices failed because the applications went through the group of patent examiners who specialize in nuclear science. But he says that by carefully structuring another application, the company was able to steer the patent to a different group of examiners, who handle electrochemistry. “Our patent attorney was very helpful in this process,” says Reding. Attempts to reach the examiners who approved the patents have been unsuccessful. A check of the USPTO Web site (http://www.uspto.gov/) also reveals other odd devices. Patent 5,830,064, for example, was granted to a company called Pear Inc. for an electronic gizmo that is meant to detect the skewing of a random signal caused by psychic forces. Pear Inc. is associated with PEAR, the Princeton Engineering Anomalies Research laboratory, which is run by Robert Jahn and Brenda Dunne, longtime parapsychology researchers who are named in the patent as co-inventors. According to the patent, the device could be used to detect the “volitional state of one or more persons” and could control games, computer displays, and appliances. Pear Inc. has been renamed Mindsong Inc. and its Web page (http://www.mindsonginc.com/) carries a press release about the patent, calling it the first patent for “devices responsive to intention of operators physically isolated from the device.” Mindsong offers a product based on this “patented technology” that is claimed to control electrical devices plugged into ac outlets on the box, for$425.00 plus shipping and postage.

At first glance, says physicist Marc Sher of the College of William and Mary in Williamsburg, Virginia, the patent seems to describe conventional technology—a circuit that generates a random signal and other electronics for detecting anomalies. “You look at the title and the abstract and it looks okay,” Sher says. “Then you look at the background information and the rest of it, and it goes off into ga-ga land.” If this kind of psychic control worked, Sher notes, it would be the “biggest thing since Galileo. But it has never been confirmed.”

Mindsong CEO John Haaland has an answer for the skeptics: “Well, they should buy it and try it.” Haaland, a former vice president of the Pillsbury food company who has a Ph.D. in biophysics from the University of Minnesota, feels that the critics have not done their homework. “They should dig harder,” he says. Haaland believes the forces at work are based on the “quantum coherence of living systems,” which skeptics do not understand, he says. What's not in doubt is the value of the patent to his company. “It's a very important piece of our portfolio,” he says. “We've been talking to investors, and the patent is a key part of our market strategy.” Haaland says the company has sold 32 of his devices so far.

The examiner for that patent, George Manuel, explained in a telephone conversation that he normally works on medical devices, but because of the backlog he was temporarily assigned to work on patents for games and toys, which is how the Pear patent is classified. Manuel said he didn't find anything outrageous about the patent. “I feel comfortable that this one was issued,” he said. “I assume that what is put forth is legitimate.”

## Trust or verify

Then there is patent 5,748,088, granted for a device to locate “entities” by “dielectrokinesis.” A product based on this patent is the LifeGuard system sold by DKL Inc. for the purpose of locating humans behind barriers. DKL says the LifeGuard can detect the electric and magnetic fields produced by a human heart at distances of up to 600 meters by means of a probe attached to a swivel mount. The LifeGuard products are currently being marketed for about $8000 each (operator training is extra) to law enforcement agencies and search-and-rescue teams for detecting intruders and locating disaster victims. DKL declined to release sales figures but said that “hundreds” of units are in use. A group at Sandia National Laboratory in Albuquerque, New Mexico, however, concluded last year that the device is ineffective, based on double-blind performance tests as well as a “teardown” and physical analysis done at the request of the National Institute of Justice.* The Sandia group also examined DKL's scientific claims. According to a company brochure, the LifeGuard's probe swivels to point to a distant human body because of “dielectrophoresis,” a term coined by University of Oklahoma physicist Herbert Pohl in the 1960s for the tendency of uncharged, highly polarizable materials to point toward the strongest part of a nonuniform electric field. As most scientists understand it, however, dielectrophoresis is a weak effect seen only in powerful electric fields. When the Sandia group ran calculations using Pohl's own equations, they concluded that “there is no possibility that the DEP [dielectrophoresis] effect is responsible for the rotation of the antenna assembly.” The executive summary of the physical analysis further concludes that the LifeGuard is not based on “dielectrophoresis or any other scientific principles as understood by the scientific and engineering community.” Conover, who is a 30-year veteran of search-and-rescue operations, says the LifeGuard is essentially a fancy dowsing rod. “It is clearly based on magical thinking and not scientific thinking.” He says many search-and-rescue teams are financially strapped. “This expense could wipe out some units,” he complains, “and it takes resources away from proven methods.” DKL President Howard Sidman says the company stands by the product and that the critics do not understand the device. “The [Sandia] report is rubbish,” says DKL chief engineer Bob VanDine. Spokesperson Nancy Wolcott says that the performance testing did not follow the operator's manual, although both the Sandia report and DKL say that one of DKL's employees was the operator for the test. DKL claims that its own analysis of the Sandia data and testing by other labs hired by the company show that the product is reliable. Further, says DKL, testing by a security company in Belgium and a crime-prevention group in Los Angeles gave 100% success rates. Wolcott claims that Sandia has a conflict of interest because the lab is trying to sell its own sensor technology. The examiner who handled this patent, Nina Tong, said that her job is to check and see if the claims are covered by previous patents. “I tried to look up ‘dielectrokinesis,’ but I couldn't find it,” she says. “I trusted them that it works as they claimed, and I assumed that people skilled in the art would use this word all the time.” Tong is an assistant examiner with a couple of years' experience at the PTO and a bachelor's degree in electrical engineering. The primary examiner who signed off on the patent, Thomas Mullen, said that he typically gives the application only a quick look to make sure all the parts are filled out. Aharonian says that the problem goes beyond inexperienced patent examiners to the pressure to process paper for what USPTO calls its “customers”—the patent applicants—which leaves little room for quality control. “The big betrayal at USPTO,” says Aharonian, “is that they forget they have two customers: the applicants, and the American people on whose behalf the applicants are granted monopoly rights.” One examiner, who requested anonymity, says that priorities have shifted at the Patent Office. “When I started several years ago, we were told ‘When in doubt, reject.’ But now, it's ‘When in doubt, issue the patent.’” Nicholas Godici, the deputy assistant commissioner for patents, refused to comment on any specific patent and denied that examiners were being hired over the phone. He added that he was satisfied with the patent examination process. Moreover, he said, the USPTO doesn't check inventions to see that they work. “We assume the information provided in an application is accurate. We don't have lab facilities or do testing, but we may ask for additional data from the inventor,” he said. Godici concedes that the public views patents as a stamp of approval but says that's a misunderstanding. Patents are nothing more than “a legal right to exclude others from using or profiting from an invention.” Yet Clean Energy's Reding says they carry an additional cachet. “We've raised$5 million from investors,” he says. “The fact that the U.S. Patent Office has declared your invention novel and unique is clearly valuable.”

11. SCIENTIFIC COMMUNITY

# A Free Energy Enthusiast Seeks Like-Minded Colleagues

1. David Voss

One patent examiner is working to make the Patent Office more hospitable to fringe energy technologies, including cold fusion: Thomas Valone. Valone, who has worked for 4 years as a patent examiner and has a master's degree in physics, is also president of a Washington, D.C.-based outfit called the Integrity Research Institute (IRI), which advertises books and videos on antigravity, mind control, and unconventional energy sources on its Web site. In an e-mail message broadcast last year on Internet news groups dealing with fringe science, Valone called for “all able-bodied free energy technologists” to “infiltrate” the Patent Office. Valone also secured government sponsorship—it was later withdrawn—for an IRI-organized conference on cold fusion, tabletop nuclear transmutation, and various other unusual energy proposals.

Valone's e-mail message offered to accept résumés at his offices at IRI and to forward applications to the appropriate supervisor at the Patent Office. Valone says he is simply trying to spread the word about free energy devices, which he feels are misunderstood. And he was briefly successful in recruiting a kindred spirit, Paul LaViolette, who was hired last year and resigned from the Patent Office on 9 April. According to the October 1998 issue of the Unofficial Gazette, a newsletter of the Patent Office employees' professional society, LaViolette's interests include assertions that antigravity technology was incorporated into the design of the B2 bomber and that the Sphinx is a 16,000-year-old cosmological cryptogram.

LaViolette confirms that Valone helped recruit him and says the Unofficial Gazette's portrait of his interests is accurate. He did not issue any patents during his short tenure, and those issued by Valone appear to be for conventional terrestrial technology. But conventional science was not the focus of the IRI's First International Conference on Free Energy. IRI persuaded the State Department last year to include its conference in the department's Open Forum program, a prestigious venue for discussions of issues in foreign policy, then promptly sent out notices in official government envelopes.

After Bob Park of the American Physical Society wrote about the meeting in his tart e-mail newsletter, What's New, the red-faced State Department insisted on having the papers peer reviewed. None of the dozen or so talks passed muster. “The papers ranged from the mediocre to the truly weird,” says a physicist at the State Department who was involved in carrying out the peer review. “Not one of them showed any understanding of modern science.” As a result, the State Department did not host the Conference on Free Energy.

But the conference was not canceled. At first, Valone moved the meeting to the Department of Commerce, where as a department employee he was able to reserve an auditorium. The meeting title was changed to the Conference on Future Energy, still hosted by IRI but advertised as being “under the auspices” of the Commerce Department. When Commerce senior staff learned of the conference, permission for use of the auditorium was withdrawn. The meeting took place at the end of April at a hotel in Bethesda, Maryland.

12. BIOLOGICAL CONTROL

# Plan to Import Exotic Beetle Drives Some Scientists Wild

1. David Malakoff

Some biologists worry that a U.S. plan to import Chinese beetles to munch on an invading tree will harm an endangered flycatcher

Next month, a handful of researchers across the western United States will receive some long-awaited express mail: cartons containing about 100 small yellow-and-black beetles. If field tests pan out, next year the Chinese leaf-eating beetles will be set free in what one scientist calls “one of the biggest and most controversial biological weed- control projects ever.” The beetles' mission is to munch on saltcedar, an Asian import that ran wild last century and now forms dense thickets along many western waterways. But although biologists agree that saltcedar is an ecological menace, many fear that the imported insect could become just as troublesome as its exotic target.

The beetle, they worry, will develop a taste for native plants as well as for saltcedar. And they note that in an ironic twist of ecology, the saltcedar now provides nesting habitat for a bird whose native habitat it destroyed: the endangered southwestern willow flycatcher. These concerns stoked nearly 5 years of fierce debate within the U.S. Fish and Wildlife Service (FWS), the agency charged with both protecting the bird and uprooting the invading plant. The debate isn't over, but last week the government approved a scaled-back plan to introduce beetle swarms into seven states.

Originally introduced as a windbreak and to control erosion, saltcedar now covers almost 500,000 hectares in 15 states. Also known as tamarisk, it has thrived in part due to its ability to sink tap roots deeper into the dry soil than native species. It also possesses remarkable fecundity: A single tree can produce more than 500,000 seeds a year. But biologists say people have aided its spread by building flood-damping dams, pumping down water tables, and allowing livestock to overgraze stream banks. “Tamarisk is wonderfully adapted to the highly modified rivers we've created,” says Rob Marshall, a biologist with the Nature Conservancy in Tucson, Arizona.

Saltcedar thickets have proved disastrous for many native species, crowding out the cottonwood trees and willows that provide food and shelter for bighorn sheep and many birds. Indeed, the plant's spread was one factor cited by the FWS for the decline of the flycatcher, a streamside resident declared endangered in 1995. Fears that saltcedar will drive more species to the edge have put the plant at the top of ecologists' hit list. But most efforts to oust saltcedar—using everything from chemicals to bulldozers—have proved fruitless or prohibitively expensive.

A dozen years ago, such problems prompted C. Jack DeLoach, a U.S. Department of Agriculture biocontrol researcher in Temple, Texas, to survey the hundreds of insects that attack saltcedar in its home range. After choosing 21 candidates, collaborators in Israel, China, and elsewhere in 1991 began to conduct “host range” studies designed to make sure the insects would not attack other plants, including valuable crops. In late 1994, government review panels approved the Chinese beetle for import and release.

The flycatcher's designation as endangered in early 1995, however, led some FWS biologists to worry that the beetles might eliminate too many nesting trees. In Arizona, for instance, nearly 90% of the state's 150 flycatcher pairs now nest in saltcedar. As a result, the FWS's southwestern office demanded further study. The controversy put the bird's advocates “in the unenviable position of defending tamarisk,” says Marshall, the agency's former lead flycatcher biologist.

But others within the agency were firmly in the beetle's corner. Their views won out late last year, when FWS headquarters overruled its southwestern outpost and announced that the beetles' likely impact on tamarisk posed no significant threat to the flycatcher—and that their presence might even benefit dozens of other threatened species by slowing the spread of the shrub.

Last week, however, the agency decided to limit that endorsement after holding several days of sometimes testy staff meetings. In a compromise expected to be approved later this month, FWS officials agreed to let DeLoach deploy special outdoor observation cages at eight study sites. That's four fewer than originally planned, and keeps the beetle out of two sensitive river valleys in Texas and New Mexico. The agency also will require DeLoach's team to pass another environmental review next year before it can release the beetles.

The scaled-back plan hasn't pleased everyone. Some biologists say that even if the beetles work, there is no guarantee that native vegetation will return to soils made too dry or salty by decades of abuse. “You could scrape saltcedar off the face of the Earth, and native stuff still wouldn't come back” unless grazing and water-management practices change, says biologist Robert Ohmart of Arizona State University in Tempe. He and other scientists also worry that the beetles could follow other introduced insects and attack nontarget plants or become deadly competition for native insects (Science, 22 August 1997, p. 1058).

But DeLoach says critics have overstated the fears of opening Pandora's box. Few of the more than 250 insects or disease- causing organisms intentionally released in 55 nations over the last century have run seriously amok, he claims—and he notes that only about half have even had any success controlling their target plants. Indeed, he originally had planned to give the beetle a boost by releasing it together with a Middle Eastern mealybug that also has a taste for tamarisk.

For the moment, DeLoach is happy to be finally moving ahead with the caged tests. But even a supporter concedes that eventually setting the beetles loose is inherently risky. After all, says ecologist Jeff Lovich of the U.S. Geological Survey in Riverdale, California, “it's not like you can call the bugs back.”

13. SCIENCE AND RELIGION

# Subjecting Belief to the Scientific Method

1. Constance Holden

A stock tycoon hopes to train scientific inquiry on the mind of God; so far, he has generated a lot of publicity

Chimpanzees squabble over mates, over food, over, well, you name it. But do they forgive their erstwhile foes—or harbor grudges? That's what primatologist Frans de Waal of the Yerkes Regional Primate Research Center in Atlanta aims to find out. He's training 20 chimps to operate joysticks that will allow them to scroll through a gallery of mug shots of their companions. De Waal hopes to open a window into the minds of his wards by observing chimp facial expressions and other behaviors before and after fights. The goal, he says, is “to see how long they remember a fight and how emotionally charged is the memory, and how the memory is affected by reconciliation.”

If you guessed this 3-year project is not sponsored by the National Science Foundation, you are correct. Bankrolling the provocative research is the John Templeton Foundation and its $10 million “campaign for forgiveness research.” Riding a wave of publicity it has received over the past year or so, the 12-year-old organization is going full throttle to yoke science to its mission of breathing new life into religion and ethics. The foundation, based in Radnor, Pennsylvania, is waging its crusade on multiple fronts: staging conferences (see sidebar), offering prizes, designing courses, churning out books, and funding research such as de Waal's as well as a program on science and religion at the American Association for the Advancement of Science (AAAS), which publishes Science. “If spiritual information doesn't begin to speed up, all religions will become obsolete,” Sir John Templeton, the foundation's prime mover and wellspring of funds, told Science. With a war chest of$800 million, Templeton has, in essence, set out to do nothing less than accelerate the evolution of religion.

The foundation's goal, in its own words, is to “encourage … worldwide explorations of the moral and spiritual dimensions of the universe and of the human potential within its ultimate purpose.” According to director Charles Harper, a planetary scientist formerly at Harvard University, scientists must play a central role in this exercise for there to be a “rapprochement” between science and religion. One prominent Templeton grantee says he has no reservations about the foundation or its philosophy. “I think they're great, and I'm a grouchy atheist,” says Stanford primate researcher Robert Sapolsky. “I have detected no hidden agenda in the way they function.”

To some critics, however, Templeton's agenda is futile, if not deceptive. “All this activity creates a flurry of illusion that science and religion are finding common ground. In fact, I think nothing in particular is happening,” says physicist Stephen Weinberg, an atheist who has debated religious scientists at Templeton-sponsored symposia. “They want to have a kind of friendly reunion of scientific and religious intellectuals, which can reestablish religion as something with a higher degree of intellectual respectability than it has among most scientists,” says Weinberg, who opposes this goal.

Other scientists have mixed feelings about the Templeton-led science-and-religion mind meld. Biologist Ursula Goodenough of Washington University in St. Louis, for one, complains that the dialogue the foundation promotes sometimes appears as an effort to prove that God exists. She points to an essay contest on “expanding humanity's vision of God,” which, she notes, encourages contestants to muster scientific evidence for a creator. But Goodenough says she approves of many other foundation activities that “really do seem open-minded and genuinely devoted to addressing real questions that aren't being addressed in any other forum.”

View this table:

Before embarking on his crusade to get scientists and theologians talking to each other, the Tennessee-born Templeton, now 86, amassed a fortune as an investment counselor and Wall Street wizard who founded the Templeton Growth Fund. In 1972 the former Rhodes scholar established the annual Templeton Prize for Progress in Religion, worth $1.3 million this year, awarded to someone deemed to have shown “extraordinary originality” in advancing understanding of spirituality. Thinking even bigger, in 1987 Templeton, a knighted British citizen, started the Templeton Foundation, which he oversees from the Bahamas with the help of a global network of advisers—mostly scientists, theologians, and business leaders—that includes former Treasury Secretary William Simon and historian Gertrude Himmelfarb, professor emeritus at City University of New York. Although a few of the several dozen advisers are Moslem, most are Christian. But that doesn't stop Templeton, a Presbyterian and self-proclaimed “enthusiastic Christian,” from prescribing an antidote for what he sees as ailing religion in general. Whereas science has progressed rapidly, religions “resist … new concepts,” Templeton says. He believes religion would benefit by following the scientific model. After all, he says, “scientists who study total world information say it doubles every 3 years. Chances are that spiritual information”—insights into how to forgive or be virtuous, for example—“has only doubled since the time of Christ.” Templeton rarely passes up an opportunity to disseminate such information, through books and “law of life” aphorisms posted daily at the foundation's Web site (sample: “Reverse the word evil and you have the word live.”) As part of his effort to reach out to the scientific community, Templeton has built a close relationship with the 18-year-old Center for Theology and the Natural Sciences at the University of California, Berkeley. Last year the foundation gave the center$12 million to run a program aimed at developing academic courses on science and religion. Another Templeton-backed center activity was a conference last June on “Science and the Spiritual Quest.” Some people complained that the roster was stacked, as all scientist-speakers were believers; organizers say that was just the point. The conference catapulted the subject to national prominence—Newsweek ran a cover story entitled “Have Scientists Discovered God?” Indeed, anywhere there's a God-and-science splash, Templeton is likely to be behind the scenes. Last fall's PBS program on “Faith and Reason,” for example, counted Templeton as a major sponsor.

As the foundation enjoys a rising profile, it is making friends among scientists. Last year it doled out $40 million, more than half of which went to “scientists who are trying to use science as a method to discover more spiritual information,” Templeton says. A centerpiece of this effort is a program that sponsors 60 projects investigating forgiveness and reconciliation in animal populations, families, tribes, and other groups (see table). Another thrust is “spirituality and healing,” including a project led by Harvard's Herbert Benson on whether “intercessory prayer” helps sick people get well (Science, 18 April 1997, p. 357). Some foundation-sponsored work has started to find its way into peer-reviewed journals: For example, a Duke University team reported last October in the Southern Medical Journal that frequent churchgoers had fewer and shorter hospital stays than nonchurchgoers. Some scientists contend that all this research is generating more hot air than light and belittle Templeton's quest for evidence of a creator. It's “not only unscientific, it is antiscientific” to posit any force other than the laws of physics behind the universe's creation, says Robert Park of the American Physical Society. Others are concerned over how the foundation picks conference speakers. Goodenough cites a lack of balance, she says, in a Templeton-supported lecture series, which at Gonzaga University in Spokane, Washington, last spring featured science historian Stephen Meyer, a prominent advocate of “intelligent design.” That term, says Harper, is associated with “neocreationists”—Christians who neither defend the biblical account of creation nor fully embrace evolution. Harper, a Christian who says he believes in a “purposive order” to the universe, defends giving Meyer a platform. Although Harper says he and his colleagues at the Templeton Foundation don't buy intelligent design, they also don't believe in “blacklisting scholars based on their points of view.” When it comes to research, no one seems to have qualms over how the foundation spends its money—least of all the recipients. A symposium last month on “the biology of belief and trust” “wouldn't have been supported” if the foundation did not exist, says organizer Randolph Nesse, an evolutionary psychologist at the University of Michigan, Ann Arbor, the symposium's venue. The meeting explored how “relationships based on trust” evolve, says Nesse, which is related to how societies adopt moral codes and religion. It's all right by him if the foundation wants everyone to believe God created the universe, Nesse says—Templeton didn't ask him to purvey that message when it gave him$20,000 for the symposium.

University of California, Irvine, biologist Francisco Ayala, an ordained priest who is a foundation adviser as well as head of the AAAS religion program's advisory board, backs what he calls Templeton's “idealistic” goal of “understanding God and spirituality through science.” Templeton says he expects “100-fold more spiritual information within a century or two”—a goal Ayala calls “naïve.” After all, accumulating religious insights may be a bit harder than growing a mutual fund—barring a miracle, that is.

14. SCIENCE AND RELIGION

# Searching for Answers to Cosmic Questions

1. Constance Holden

The laws of nature are “cold and impersonal,” pronounces physicist Steven Weinberg. The world is “shot through with signs of mind,” counters physicist John Polkinghorne. Debate of this kind is not standard fare for your average scientific conference—unless it's one sponsored by the John Templeton Foundation (see main text). The organization has trotted out this polarized pair—atheist Weinberg, a professor at the University of Texas, Austin, and Anglican priest Polkinghorne, former president of Queens College at Cambridge University—at several fora across the country, including “Cosmic Questions,” an event held last month at the Smithsonian Institution in Washington, D.C. and co-sponsored by the American Association for the Advancement of Science (AAAS).

Weinberg versus Polkinghorne makes for good theater, but the conference had a deeper purpose. The idea was to bring scientists and theologians together to air their views on three mind-numbing questions: “Did the universe have a beginning?” “Was the universe designed?” and “Are we alone?” Some scientists say the time is ripe to seek out common ground. “A new scientific cosmology is emerging today,” said physicist Joel Primack of the University of California, Santa Cruz, who illustrated this notion with a thrilling computerized plunge through the Hubble Space Telescope's deep field, carrying viewers back 15 billion years, nearly to the birth of our universe. Indeed, said physicist Robert John Russell, a minister of the United Church of Christ and founder of the Center for Theology and the Natural Sciences at the University of California, Berkeley, “it's very important [for theologians] to take the big bang seriously.”

At the conference, participants circled around a concept providing a Venn diagram-like overlap between science and religion. Called the “anthropic principle,” in its simplest form it states, in the words of physicist Frank Tipler of Tulane University in New Orleans, that “the observed properties of the universe are consistent with human life evolving in it.” That is, if the basic physical constants—such as gravity or an electron's mass—were the slightest bit off, life could not exist. If gravity were stronger, matter would have collapsed in on itself; weaker, and matter would have pulled apart too fast to coalesce. Both the universe's “intelligibility” (it follows laws of science) and its “suitability” for life are evidence of the hand of God, argued Anna Case-Winters, a professor of theology at McCormick Theological Seminary in Chicago.

More than a few scientists, however, dismiss the principle: You can't talk about odds-defying circumstances when you have a sample of only one universe, noted physicist Alan Guth of the Massachusetts Institute of Technology. All you can say is “If it did happen, it could happen.” Indeed, some scientists think science has little to gain from such “Cosmic Questions” exercises. According to paleobiologist Stephen Stanley of Johns Hopkins University, applying theology to ethical debates in science—one obvious mechanism for bringing about a science-religion encounter—“will simply complicate an already complex issue.”

Organizers see things differently. “The major accomplishment of the conference was to contribute to greater understanding across disciplinary boundaries,” says Audrey Chapman, who heads the AAAS's 3-year-old Dialogue Between Science, Ethics, and Religion, which has a 5-year, $1.3 million grant from the Templeton Foundation. A dialogue, at least, can't hurt, says Guth. “Much of the brainpower that has been thrown at ethical questions in science has come from theologians,” he says, “so it is good for scientists to stay in touch.” 15. NUCLEAR PROLIFERATION # U.S. Sanctions Block People But Not Goods From India 1. Pallava Bagla* 1. With reporting by Jeffrey Mervis in Washington. A U.S. ban on exports of sensitive weapons material also prevents some Indian scientists from visiting some U.S. civilian labs MUMBAI AND WASHINGTON, D.C.—One year after India conducted nuclear tests that triggered new U.S. sanctions against dozens of research institutions affiliated with the country's defense sector, it's business as usual for a few U.S. companies buying products and other technology from blacklisted institutions. The deals—covering equipment to monitor power plants and technology to interact with communications satellites—are possible because the sanctions prohibit U.S. exports, but not imports, of technology. Although hailed by Indian officials as proof of their scientific prowess, the transactions are conducted under rules that scientists from both countries say often defy logic. Those who developed the technology can't travel to the United States to install the equipment, for example, and anything that needs to be repaired can't be sent back because it would violate U.S. export restrictions. And the one-way technology flow also threatens long-standing civilian collaborations that scientists say have nothing to do with nuclear weapons. The latest U.S. sanctions, imposed last summer, are meant to deprive India of material and knowledge that might advance its nuclear weapons or ballistic missile program. (The sanctions also apply to Pakistan, which conducted its own tests last year in response to India's actions, but there are fewer U.S.-Pakistani scientific interactions.) At their core is a ban on visits to the United States by scientists working at more than 60 institutions, including India's civilian nuclear centers, fundamental science institutes, and its space agency. Visits by U.S. scientists to these institutes are also barred. “We want to limit the threat to U.S. security” from foreign scientists who might bring home information gleaned from their visits, says a spokesperson for the Commerce Department, which enforces the sanctions. “We don't mind learning their secrets, but we don't want to share ours.” U.S. officials say the sanctions could be lifted if India shows a greater commitment to arms control by ratifying the Comprehensive Test Ban Treaty and an agreement on the control of fissile materials. In the meantime, the Department of Energy (DOE) is pressing the State Department for an immediate change in its current policy. Instead of issuing a blanket denial on visa requests by scientists from a blacklisted institution, say DOE officials, immigration officers should examine each request on its merits. The proposal, under review by Energy Secretary Bill Richardson, would also need State Department approval before going out as a policy directive to embassy staffs. Such a change in policy might yield an immediate payoff for an international collaboration of high-energy physicists working at the Fermi National Accelerator Laboratory (Fermilab) in Batavia, Illinois. The team—450 scientists from 13 countries, including India—is spending$40 million to upgrade a five-story-high detector, called D0, for an experiment due to start late next year on the Tevatron, the world's most powerful proton accelerator. In March 1998, Indian scientists from the Tata Institute of Fundamental Research (TIFR) in Mumbai shipped their contribution, a $500,000 scintillation counter and shield to block out cosmic rays from the outside and to identify muon particles generated in the collision of experimental particles. The D0 team is beginning to assemble and test the various parts of the massive detector. Under normal circumstances, that would prompt a visit by four or five senior scientists from TIFR to oversee the installation, along with a handful of students in accelerator physics eager to learn the tricks of the trade. But the Indian component remains packed up in a crate: TIFR is on the government's blacklist, and none of the team is able to obtain a U.S. visa to work at Fermilab. “We'd like them to come now, but they've been told that's not possible,” says David Cutts, a physicist at Brown University who chairs the institutional board for the experiment. “Installing the equipment is more than just tightening the nuts and bolts. We need good experimentalists to make sure that it's working properly. And they are a strong group that has been working with us for more than a decade.” The team expressed its concern in a 10 February letter to Richardson and Secretary of State Madeleine Albright. In a 30 March reply, State's country director, Gary Usrey, agreed that the sanctions have “negatively affected scientific collaboration and the search for fundamental knowledge.” But he wrote that the impact is a “regrettable consequence of India's and Pakistan's decisions to conduct nuclear tests” and that “scientific collaboration will continue to be affected” unless the two countries modify their policies. Contacted last week, Usrey explained that “there's a presumption of denial on trips by officials from those entities because they are part of India's nuclear program.” Cutts believes that such an argument is spurious. “[TIFR] is like Fermilab in what it works on—basic high-energy physics,” he says. “Of course, both labs are funded by a department of energy that is also responsible for nuclear weapons. But that's not what they do.” TIFR's director, Sudhanshu S. Jha, also objects strongly to the current policy. “It seems the U.S. is only interested in the equipment and not the brains that make it work,” he scoffs. That description could also apply to the ongoing ties between General Electric Co. (GE) and the Bhabha Atomic Research Center (BARC) in Mumbai, India's main nuclear weapons laboratory. BARC felt the lash of the sanctions last summer when its former chief, Rajagopala Chidambaram, was denied a visa to attend a crystallography meeting outside Washington, D.C. (Science, 24 July 1998, p. 494). But last month the center received its third repeat order from GE's Indian affiliate to supply the industrial giant with components that use radioactive thorium dioxide to monitor the accumulation of moisture in the electric generators of power plants. “Though the [monetary] value may be small, what really matters is that it is an American order,” says BARC's current director, Anil Kakodkar, who sees the contract as an endorsement of his lab's technical ability. The Indian Space Research Organization (ISRO), which comes under the sanctions because it is widely believed to be helping in the development of the country's missile program, still manages to do business worth millions of dollars with several U.S. companies. Only last month ISRO completed a 10-year,$110 million deal to hand over 11 transponders from the Indian communications satellite INSAT 2E to the Washington-based satellite consortium Intelsat. N. Sampath, a director at ISRO, says the arrangement demonstrates that U.S. companies “can get comparable quality at a cheaper price” from India. ISRO still manages to export $2 million worth of satellite subsystems and components to U.S. companies, he notes, although he estimates that the agency has lost as much as$3 million in sales from U.S. companies that are reluctant to tackle the additional paperwork required to deal with an institution on the sanctions list.

Despite such scientific and economic setbacks, most Indian officials say that the sanctions have not prevented the country's defense, space, and atomic energy labs from carrying out high-quality work. “We have lived under embargoes ever since 1974 [when India conducted its first nuclear test] and have learnt how best to turn it to our advantage,” says Chidambaram, now head of India's atomic energy program. “We look at the sanctions as a blessing in disguise.”

16. CELL BIOLOGY

# Nuclear Transport Protein Does Double Duty in Mitosis

1. Elizabeth Pennisi

Several studies now show that Ran, which plays a key role in nuclear transport, is also a trigger for the formation of the mitotic spindle

Like an actor who gets typecast as a villain and can't ever get recognition for his full range of talents, Ran is a protein whose multifaceted nature has often been ignored. By 3 years ago, it had gained fame for its role in shipping other molecules in and out of the cell's nucleus (Science, 20 February 1998, p. 1129), and that view of its role stuck. But now Ran's versatility is starting to be appreciated.

Researchers have identified what appears to be a trigger for cell division, and it turns out to be none other than the nuclear transport protein. Several teams, two of which report their results in this issue, have evidence indicating that Ran, when bound to the high-energy molecule guanosine triphosphate (GTP), prompts the formation of the mitotic spindle, an array of polymer threads called microtubules that help draw the chromosomes to the opposite sides of the dividing cell. Exactly how RanGTP does this is unclear, but the results signal the dawn of a new era of exploration into Ran and may help researchers resolve long-standing questions about the timing and progression of this early stage of cell division.

The new findings also bring Ran's career full circle. It originally earned a name for versatility when researchers implicated it in a puzzling array of cellular activities, such as modifying the RNA in the ribosomes, the cell's protein-making organelles, stabilizing chromosomes, and controlling the progression of cell division. But with the discovery of Ran's role in nuclear transport, they concluded that the various derangements seen in cells with a faulty Ran gene were the indirect consequence of a disruption in nuclear transport. For example, cells might not divide normally because the messenger RNA encoding a necessary signal for mitosis couldn't get out of the nucleus and thus its protein product could not be made.

Last year, however, cell biologist Takeharu Nishimoto and his colleagues at Kyushu University in Fukuoka, Japan, reported the first clue that Ran might be directly involved in controlling spindle formation. They identified a Ran-binding protein that they called RanBPM and showed that it accumulates on the centrosomes, which form inside the cell and serve as the starting points for the growing mitotic spindle. “That was the first direct link between Ran and the spindle system,” says Shelley Sazer, a cell biologist at Baylor College of Medicine in Houston.

To try to pin down this link, Nishimoto turned to a system that cell biologists have used for years to study spindle formation. It involves breaking up an unfertilized frog egg, separating out its DNA, and extracting the remaining soluble contents. By adding sperm, whose own nuclear membranes have been removed, to these extracts in the test tube, researchers can reproduce mitosis: The sperm DNA reorganizes into discrete chromosomes that line up as if they were still in an intact cell undergoing division. In addition, within minutes, asters, starlike clusters of microtubules branching out from the centrosome, also form, the first steps toward production of the spindle. Thus, researchers can study the biochemical requirements of mitosis without the complications of having intact cellular structures like the nucleus.

In the current work, which is described on page 1356, Nishimoto and his colleagues used the egg system to test the role of different forms of Ran in spindle formation. Ran belongs to a family of enzymes called GTPases, which cycle between a high-energy form with GTP attached to it and a low-energy form carrying guanosine diphosphate (GDP). In the nucleus, a protein called RCC1 keeps the concentration of the GTP form high by converting any RanGDP entering the nucleus to RanGTP. And the work showed that the GTP form of Ran, which would be released from the nucleus when the nuclear membrane breaks down at the beginning of mitosis, is critical for the aster formation. For example, when the researchers added antibodies that block the activity of RCC1, they saw very few asters in the egg extract system. But when they added back large amounts of RanGTP, aster formation was nearly normal.

In a similar vein, Andrew Wilde and Yixian Zheng of the Baltimore, Maryland, laboratory of the Carnegie Institution of Washington linked RanGTP to the dramatic remodeling of microtubules that goes on as cells enter mitosis and ultimately form the spindle. In nondividing cells, these threads, which serve as the cell's internal transport system, are long and fairly permanent. In preparation for mitosis, they suddenly disintegrate. In their place, short fibers appear, and the microtubules “become quite dynamic,” continually forming, breaking down, and re-forming, Zheng explains. “It was clear that the microtubules had to be regulated.” Then as the nuclear envelope breaks down, the spindle begins to take shape.

Zheng and Wilde suspected that a GTPase might be involved in microtubule regulation, as another GTPase regulates the polymerization and breakdown of a different fiber-forming protein, actin. But they didn't seriously consider Ran until last year, when Nishimoto described finding RanBPM on centrosomes. “That seemed to make the link [to] Ran,” Zheng recalls.

Using the egg extract system, Zheng and Wilde found that it took just a little RanGTP to prompt spindle formation. “It was such a surprise to see such a dramatic effect when we put in just a single protein,” Wilde says.

A third group that has now linked Ran and its proteins to spindle formation includes Mary Dasso, Petr Kalab, and Robert Pu of the National Institute of Child Health and Human Development (NICHD) in Bethesda, Maryland. In addition, their work suggests that proper progression through the cell division cycle requires several Ran-related molecules, including one called RanBP1, which helps convert RanGTP to RanGDP in the cytoplasm.

That conclusion is further buttressed by results reported last year in the Journal of Cell Science by Patria Lavia, a cell biologist at the University of Rome “La Sapienza,” and her colleagues. When Lavia's group forced cells to make RanBP1 all the time, the cells either failed to divide properly or stopped dividing altogether. The right amount of RanBP1 seems “to be required for proper [microtubule] dynamics,” says Lavia.

But because those effects were all in intact cells, they could have been the result of disrupted nuclear transport. In their latest series of experiments, described in the May issue of Current Biology, the NICHD researchers found that if they added RanBP1 by itself to the egg extract system, the spindles didn't form, presumably because the protein converted the RanGTP in the samples to RanGDP. But if they then added RCC1, it neutralized the effects of RanBP1, and asters again appeared. “It's as if you have to regulate the ratio of [Ran]GTP to [Ran]GDP in some way,” notes Mark Rush, a cell biologist at New York University Medical Center.

The picture developing from these results shows that “Ran could be the signal that acts in mitosis to stimulate microtubule polymerization or stabilization,” Zheng says. But, she adds, the amount of RanGTP must be tightly regulated for the spindle to form properly.

Still, skeptics point out that researchers haven't yet shown a direct connection to the proteins involved in spindle formation. “It would be nice to see that [RanGTP] binds to a purified protein,” says Baylor's Mary Moore. Unless that can be shown, Pamela Silver, a cell biologist at the Dana-Farber Cancer Institute in Boston, says that the effects attributed to RanGTP in the egg extracts might be due to something else. She suggests, for example, that the egg extracts could contain key microtubule-promoting proteins that remain stuck to their transport molecules even though there's no nuclear membrane left.

For the link between Ran and spindle formation to hold up, says Sazer, “the next step will be to try to understand how the Ran system operates in vivo.” Hints are already emerging. Lavia has found that the expression of the gene for RanBP1 is normally turned on by proteins known to regulate the activity of other genes in the cell cycle. This helps ensure that RanBP1 accumulates at the right moment in the cell cycle to help control mitosis. These discoveries, and others that are rumored to be coming out soon from other labs, are making researchers like Dasso confident that Ran's multiple roles are real.

17. MICROBES, IMMUNITY, AND DISEASE

# A Symphony of Bacterial Voices

1. Evelyn Strauss

Bacteria can communicate with members of both their species and others, thereby allowing them to coordinate their activities

Hum by yourself in an auditorium and the room will swallow the sound. But wait until the rest of the chorus shows up, and together you can fill the hall with music. Some activities require a group to make an impact. Indeed, even microbes behave as if they understand that there's power in numbers.

Over the past 5 years, scientists have found that a wide variety of bacteria gather a crowd before acting in unison—a phenomenon that's been dubbed quorum sensing. “Bacteria want to do some things when they're alone and others when they're in a community,” says Bonnie Bassler, a microbial geneticist at Princeton University. “Sensing and responding to high population density allows them to switch to a new set of tasks.”

In some cases, doing so benefits not just the microbes but also the creatures they inhabit. The classic example is light emission by the luminescent bacterium Vibrio fischeri, which lives in a specialized organ in host animals such as the Hawaiian bobtail squid. The light that is turned on when the organ is sufficiently filled with bacteria eliminates the squid's shadow as it swims in the moonlight, making it hard for its predators to see it. Thus, the animals live longer, and the bacteria get a home.

In many other cases, though, the results of quorum sensing render bacteria harmful to their hosts. Some bacteria, when their population is large enough, congregate in slimy mats called biofilms, which have been implicated as a cause of many human infections (see Review by Costerton, Stewart, and Greenberg on p. 1318); others produce virulence factors, proteins that allow pathogenic bacteria to exploit their host (see Review by Galán and Collmer on p. 1322).

As researchers tease out the molecular details of how bacteria sense and respond to their neighbors, they may learn how to foul up the counting systems of specific strains. This could lead to ways to thwart the ability of pathogenic bacteria to cause illness without generating side effects by wiping out indigenous flora, as current broad-spectrum antibiotics usually do. “Instead of sterilizing the human host, you're trying to make it unfit for the particular bacteria to survive,” says microbiologist Peter Greenberg of the University of Iowa, Iowa City. Conversely, research on quorum sensing could spawn methods to enhance desired activities, such as the production by plant bacteria of antibiotics that protect their host plants from disease.

Because recent research suggests that bacteria can take a census of other species as well as their own, understanding quorum sensing will likely also reveal insights into microbial ecology. “In most natural environments, bacteria live in complex mixed communities,” says Greenberg. “Understanding how these develop and organize is going to be critical in attempts to control and manipulate them.”

## The light at the end of the tunnel

Microbiologists got their first inklings that bacteria might be counting their neighbors in the late 1960s when J. Woodland Hastings, currently at Harvard University, and his colleagues noticed a perplexing pattern of bioluminescence in V. fischeri and various other marine bacteria. During early growth, these bacteria remained dim. But when their cells reached a certain density, their luminescence shot up.

Kenneth Nealson and Terry Platt, working with Hastings at Harvard, soon traced the effect to some unknown signaling substance that the bacteria secrete into their growth medium. But that substance remained elusive until the early 1980s. Then, using as an indicator a natural V. fischeri strain that can't make the signal but does respond to it, Anatol Eberhard of Ithaca College in New York and his colleagues purified the bacterium's quorum-sensing signal, a small molecule called an acylated homoserine lactone (acyl-HSL).

Over the next several years, researchers learned that V. fischeri's quorum-sensing system has two key components: an enzyme that makes the acyl-HSL plus a protein that detects the chemical and responds to it by activating specific genes. As the bacteria multiply within a confined space, the enzyme, called LuxI, produces ever-increasing amounts of acyl-HSL. Eventually the chemical, which diffuses freely into and out of the bacterial cells, reaches a critical concentration at which it activates the second component, a protein called LuxR. The activated LuxR in turn binds to specific stretches of regulatory DNA adjacent to the V. fischeri genes needed for bioluminescence and turns them on. This system ensures that the bacteria produce light only when large numbers are present, keeping them from wasting energy when their population is too small to emit a visible glow.

By the mid-1990s, microbiologists had found that many other bacteria that are, like V. fischeri, members of the gram-negative group use the same general acyl-HSL scheme for quorum sensing. (Gram-negative bacteria are so called because they don't retain the purple Gram stain.) And bacteria use other signals as well. “It was originally thought that this weird communication system found in fish light organs was just some kind of fluky thing,” says Gary Dunny, a microbiologist at the University of Minnesota, Minneapolis. “But it turns out that it's probably quite the norm for bacteria.”

About a dozen human pathogens are among the bacteria so far known to possess an acyl-HSL quorum-sensing system. Take, for example, Pseudomonas aeruginosa, a common cause of hospital-acquired infections and of lung infections in people with cystic fibrosis. P. aeruginosa is hard to eradicate because it forms biofilms, which shelter the bacteria from antibiotics, detergents, and the host's immune system. Greenberg's team has shown that acyl-HSL production by the bacteria is what triggers their congregation in biofilms.

In addition to infecting humans, many gram-negative bacteria live on plants and in the soil, where they have such beneficial effects as protecting plants from infections or helping fix nitrogen, as well as harmful ones such as causing wilting diseases or crown gall tumors. Some of these bacteria, too, use acyl-HSLs for quorum sensing. In work described in the November 1998 issue of Molecular Plant-Microbe Interactions, for example, microbiologist Stephen Farrand of the University of Illinois, Urbana-Champaign, and his colleagues surveyed a total of 106 bacterial strains, representing seven different genera that live on various plants.

The researchers found that about half, representing all seven genera, produce acyl-HSLs and have the biochemical machinery to activate gene expression in response to acyl-HSL signals. They haven't yet demonstrated directly that the acyl-HSLs activate the gene expression or that the bacteria use them for quorum sensing. But that is likely, Farrand says: “If the bacterium goes to the trouble of making an acyl-HSL, my guess is it has—or had [sometime in its evolutionary past]—a quorum-sensing system that uses it as a signaling molecule.”

The discovery that multiple strains of plant bacteria make similar signaling molecules raised the possibility that the bacteria can hear members of other species as well as their own. To find out if such eavesdropping occurs, Leland S. Pierson III, a microbiologist at the University of Arizona, Tucson, and his colleagues turned to a plant-associated bacterium, P. aureofaciens. This microbe uses its own acyl-HSL quorum-sensing signal to determine when to make an antibiotic that helps it suppress its microbial competitors and also provides a service for its wheat plant hosts—protecting them against the fungal “take-all” disease.

In work reported in last November's Molecular Plant-Microbe Interactions, Pierson and his colleagues showed that production of the antibiotic, called phenazine, can also be turned on by signals from dozens of other bacterial strains. They isolated 700 strains of bacteria from wheat roots, let them grow in culture, and exposed a P. aureofaciens strain that doesn't itself make an acyl-HSL to the culture fluid in the lab. In 8% of the cases, signals in the fluid—which the group has since shown to be acyl-HSLs—stimulated the P. aureofaciens to make phenazine.

Pierson then went on to show that the same kind of cross talk can take place in the wild. His group prepared five sets of seeds, each coated with one of five different bacterial strains that had activated the phenazine biosynthetic gene in the lab plus a strain of P. aureofaciens equipped with a “reporter” gene that would produce an easily detectable product—it raises the freezing temperature of water—in response to acyl-HSLs. The researchers then planted the seeds and 10 days later looked to see whether the reporter bacteria on the seedling roots had reacted to signals from the five test strains. They found that all had produced positive effects. “There's a potential for communication between unrelated populations of bacteria in the real world,” says Pierson.

One reason bacteria might want to intercept their neighbors' messages, he says, is that “a large amount of signal suggests that other bacteria are growing and happy. That tells the bacteria that this is a great place to be.” Another is competition. As Pierson puts it, “a plant root is not Club Med. There are limited nutrients, and a bacterium needs to know who else is there so it can make decisions about how to expend energy and succeed in that environment.” In P. aureofaciens's case, this means making antibiotics to inhibit the growth of competing organisms.

With all this chatter, it might be hard for one species to hear its own members count. Work by Princeton's Bassler and Michael Silverman of the now defunct Agouron Institute in La Jolla, California, suggests that one way bacteria solve this problem is by using multiple quorum-sensing systems. The researchers have shown that the luminescent genes of the marine bacterium Vibrio harveyi are regulated by an elaborate mechanism that involves two discrete quorum-sensing systems. One of these uses an acyl-HSL signal, although both its synthesis and the responding molecule differ from those of more conventional acyl-HSL quorum-sensing systems. But the second uses a different, as yet unidentified signal.

Activation of either system is enough to turn on the light-producing genes, so Bassler wondered why there should be such redundancy. Experiments she has since performed with Iowa's Greenberg provided an answer. They found that a strain of V. harveyi that can sense the acyl-HSL but not the second signaling substance turns on light production only in response to signals from V. harveyi and a few other bacteria. Conversely, a strain of V. harveyi that can't sense the acyl-HSL of quorum system one but can pick up the second type of signal responds to a broad range of bacteria. “I think V. harveyi has system one for intraspecies communication and system two for interspecies communication,” says Bassler. “Maybe the combination provides it with a way to know its proportion of the total bacterial population. It could then adapt appropriately by turning on different sets of genes, depending on whether it's alone or in a consortium.”

## E. coli and S. typhimurium catch up

The discovery that V. harveyi employs unconventional molecules—non-acyl-HSL signalers and non-LuxI/LuxR proteins for making and responding to the signaling compounds—may also help resolve a paradox: Escherichia coli, a gram-negative bacterium, has served as a prototype for understanding much about bacterial life, but even though quorum-sensing systems have been turning up all over the place, this organism didn't appear to have one—at least not one involving acyl-HSLs. New findings from Bassler's group now suggest that E. coli and its close relative Salmonella typhimurium may have quorum-sensing systems similar to V. harveyi's second system.

Last year, the researchers showed that E. coli and S. typhimurium make a substance that activates luminescence gene expression in a V. harveyi mutant that can sense signals only through its second system. Since then, Bassler's team identified a gene involved in producing the substance in V. harveyi, E. coli, and S. typhimurium. [The results are in the 16 February issue of Proceedings of the National Academy of Sciences (PNAS).] Similar genes—and presumably similar signaling systems—seem to be widespread. By surveying the databases, Bassler and her colleagues have discovered related genes in over a dozen other species of bacteria.

The signaling substance this gene helps produce may tell E. coli to go on the attack once its population reaches a critical level. Working with the O157:H7 strain of E. coli, which has gained a justifiably bad reputation as a dangerous food-borne pathogen, James Kaper's team at the University of Maryland School of Medicine in Baltimore has shown that the gene Bassler identified produces a substance that induces the expression of this organism's type III secretion system. This system produces a “molecular syringe,” which injects bacterial factors critical for successful infection into animal host cells.

And E. coli's linguistic abilities may not stop there. In work reported in the 13 April issue of PNAS, microbiologist Philip Rather of Case Western Reserve University in Cleveland and his colleagues have uncovered what appears to be another quorum-sensing system in E. coli.

Like other quorum-sensing bacteria, E. coli may use its multiple systems to sense both its own numbers and those of other species. These pathogens might use other microbial “noise” to figure out when they've reached their destination. “Maybe they're picking up signals from normal bacteria in the intestine,” says Kaper. “They then recognize it as an environment in which they want to make this type III secretion system.”

Still more quorum-sensing languages have turned up in the gram-positive bacteria, which do take up the Gram stain. “We're beginning to discover that bacteria have a whole language, and acyl-homoserine lactone is just one word. E. coli has other words it can use. And gram-positives use another dialect,” Greenberg says. One case in point is Staphylococcus aureus, which causes infections ranging from small skin abscesses to life-threatening conditions such as endocarditis and toxic shock.

When this pathogen infects its hosts, it activates a set of genes that encode a variety of protein-degrading enzymes and toxins that help it disseminate in the host and wreak havoc on the immune system. There has been some disagreement about the identity of the signaling molecule that activates these genes, but the latest evidence, published in the 15 February issue of PNAS by microbiologist Richard Novick's group at New York University Medical Center in Manhattan, shows that a short peptide can do the job.

In that work, the Novick team synthesized peptides they had previously implicated in quorum sensing by S. aureus and showed that they activate the microbe's virulence genes. The fact that the researchers could duplicate the effect with completely synthetic peptides confirms that the molecules are the quorum-sensing signals, Novick says.

The Novick team has also found that although the peptides from different S. aureus strains share basic chemical features, they divide into four groups based on variations in amino acid sequences and length. And because peptides from one group can inhibit virulence-gene expression in S. aureus strains of another, the finding opens the door to developing new drugs for treating S. aureus infections.

Indeed, exploiting such dialects may turn out to be a broader strategy for turning bacteria's quorum-sensing systems against them. Researchers have found that the acyl-HSL signals of gram-negative bacteria have side chains that often vary from one bacterium to another, and some acyl-HSLs can inhibit the activity of the LuxR-like proteins of other species.

New results from Greenberg's team suggest another potential way of inhibiting quorum-sensing signals in these organisms. The enzymes that synthesize the acyl-HSL signals use a substrate that organisms from bacteria to humans exploit to construct a variety of molecules. As a result, trying to inhibit those enzymes might seem an unpromising strategy for shutting down a quorum-sensing system. But Greenberg and his colleagues found a specific molecular feature that distinguishes the acyl-HSL-synthesizing enzyme in P. aeruginosa, the biofilm-forming pathogen, as the group reports in the 13 April issue of PNAS. “This leads to the idea that you could find analogs that don't bind to proteins in humans but do bind specifically to this family of [microbial] proteins,” says Greenberg. Such compounds might thwart the production of the signaling molecule and thus of P. aeruginosa biofilms.

Techniques for scrambling or shutting down quorum-sensing systems could be valuable against some nonbacterial pathogens, too. Quorum-sensing systems are now turning up in more complex, nucleated cells. “Often ideas are discovered in bacteria and if they're great, eukaryotes use them too,” says Greenberg. Next week at the American Society for Microbiology meeting, for example, William Goldman, a microbiologist at Washington University School of Medicine in St. Louis, will present evidence that the fungus Histoplasma capsulatum, which causes histoplasmosis, a flulike respiratory disease in humans, can sense its own numbers.

Goldman and his colleagues have found that the fungal cells in dense cultures are adorned with a particular sugar whose presence correlates with the virulence of the fungus. But when they diluted those cells in fresh culture medium, newly budding cells did not display the sugar. “It continues to stay off until they reach a certain cell density and then—wham—everyone turns it on,” says Goldman. If, however, the researchers add fluid from dense cultures to the diluted fungi, the buds express the sugar right away. “Here's a trait that's clearly related to virulence, and it's turning on and off in response to cell density,” he adds.

At the same time that researchers are nailing down the biochemical details of the quorum-sensing systems, the many new findings are opening unexplored areas. Given bacteria's knack for languages, Greenberg wants to know, for example, whether they speak to animals or plants, and if so, whether the microbe or the host benefits. If the host can understand or even hear the bacteria, it might “get a leg up in responding to the impending infection,” he says. Alternatively, the bacterium may trick the host into responding inappropriately.

And now that it seems clear that different bacterial species can hear each other, “we need to figure out the role of this cross communication in the normal ecology of the strains,” says Arizona's Pierson. For example, he says, it might help the organisms coordinate tasks, with one strain producing an enzyme that inactivates an antibiotic while another secretes a substance that prevents the community from drying out. “Different species of bacteria have different capabilities, but they work together,” says Bassler. “They're behaving like multicellular organisms. That's really smart”—but perhaps no smarter than might be expected of organisms that orchestrate their activities by performing in concert.

18. MICROBES, IMMUNITY, AND DISEASE

# Is It Time to Uproot the Tree of Life?

1. Elizabeth Pennisi

More genomes have only further blurred the branching pattern of the tree of life. Some blame shanghaied genes; others say the tree is wrong

A year ago, biologists looking over newly sequenced genomes from more than a dozen microorganisms thought these data might support the accepted plot lines of life's early history. But what they saw confounded them. Comparisons of the genomes then available not only didn't clarify the picture of how life's major groupings evolved, they confused it (Science, 1 May 1998, p. 672). And now, with an additional eight microbial sequences in hand, the situation has gotten even more confusing—so confusing that some biologists are ready to replace what has become the standard history with something new.

Many evolutionary biologists had thought they could roughly see the beginnings of life's three kingdoms: the ordinary bacteria; the Archaea, which are microbes best known for living in extreme environments; and the eukaryotes, which are all organisms, from yeast to people, whose cells have distinct nuclei. Comparisons of the genes encoding microbial species' ribosomal RNAs—the RNAs that make up the ribosomes, the small cellular structures where proteins are synthesized—suggested that life began with some sort of primitive bacteria. These then diverged into two branches, one leading to the modern bacteria and the other producing the Archaea and later branching again to produce the eukaryotes. When full DNA sequences opened the way to comparing other kinds of genes, researchers expected that they would simply add detail to this tree. But “nothing could be further from the truth,” says Claire Fraser, head of The Institute for Genomic Research (TIGR) in Rockville, Maryland.

Instead, the comparisons have yielded many versions of the tree of life that differ from the rRNA tree and conflict with each other as well. Some put Archaea and bacteria together; others divide the Archaea into multiple groups. And microbiologists' response to this confusion is equally varied, as this year's Microbial Genomes III meeting, held in early February in Chantilly, Virginia, showed.

Some, pointing to evidence that microbes have swapped genes wantonly over evolutionary history, say that many of these genes are an unreliable guide to evolutionary history and the old RNA-based tree is still basically sound. But others think that it's time to uproot the old tree and are proposing candidates for new trees based on specific features of the genome and cell structure. And still others worry that gene swapping has turned the tree of life into a tangled briar whose lineages will be next to impossible to discern. “There's so much lateral transfer that even the concept of the tree is debatable,” says André Goffeau, a geneticist at the Université Catholique de Louvain in Louvain la Neuve, Belgium.

## A study in confusion

As a case study in how new data are muddying the evolutionary picture, consider Thermotoga, a heat-loving microbe discovered 13 years ago on Volcano Island off Italy. The sequence of one of the microbe's rRNAs puts it on the bacterial branch of the evolutionary tree, just a bit higher than an ancient bacterium called Aquifex, sequenced last year by Diversa. But after TIGR scientists completed Thermotoga's genome 9 months ago, TIGR's Karen Nelson and her colleagues compared gene sequences from the two microbes. They found, she says, that “there's no consistent picture [of] where these two organisms fall.”

As she described at the meeting, Nelson first identified 33 genes that are found in both Thermotoga and Aquifex, as well as in an additional 10 bacterial species, four Archaea, and the eukaryote yeast. She then used the base changes in the genes of the various organisms to construct separate trees reflecting the evolutionary history of each gene. “We could find only three situations that supported the branching order [derived] from the ribosomal subunit,” she said. “It was impossible to say whether Aquifex or Thermotoga was more ancient.”

Some evolutionary specialists suspect that such confusion is the result of rampant gene swapping. A gene acquired that way by one microbe would then look very similar to its counterpart in the donor organism, indicating their close kinship, while another gene could be very different in the two, possibly because it came from still another species. Thus, the histories inferred from the two genes would be contradictory, making the true history of the microbes difficult to discern.

This phenomenon is becoming more apparent with each new genome. The Thermotoga genome, for example, confirmed a suggestion made last year by Frank Robb and Dennis Maeder, microbiologists at the University of Maryland Center of Marine Biotechnology in Baltimore. Based on the partial Thermotoga genome, they proposed that the microbe has many genes in common with Archaea, and now that idea has been borne out. Because the two aren't supposed to be closely related, the likely explanation is that one somehow acquired genes from another, possibly because those genes were once part of mobile genetic elements capable of inserting into foreign genomes. Other researchers recently buttressed the case for such microbe-to-microbe transfers by finding what appears to be evidence for a recent transfer—evolutionarily speaking—within the past 100 to 1000 years (see sidebar).

Indeed, microbes can apparently even appropriate genes from “higher” organisms. At the microbial genome meeting, for example, Kira Makarova of the Uniformed Services University of the Health Sciences in Bethesda, Maryland, and her colleagues reported that the genome of the bacterium Deinococcus radiodurans contains several genes previously found only in plants. Thus it appears that these genes somehow got transferred to the microbe. And evolutionary biologist Winston Hide of the University of the Western Cape in Bellville, South Africa, reported that his team found that Mycobacterium tuberculosis, which causes tuberculosis, has taken on at least eight human genes. This is apparently beneficial to M. tuberculosis, as the genes encode proteins that help break down the hydrogen peroxide that some white blood cells use to kill foreign invaders. They could thus help the bacteria fight off host defenses.

Experts differ, however, on what such gene transfers mean to researchers trying to trace evolutionary relationships. Some, such as Goffeau, wonder whether they make tree-building difficult, if not impossible. These researchers point out that the transfers may have gone on even more extensively in life's early days, when the molecular machinery for replicating and processing genetic material was still universal and organisms had not yet evolved ways of getting rid of foreign DNA. If so, tree building may be a meaningless exercise. The genomes of modern microbes may be mosaics of genes from different organisms rather than descendants of any single early form of life, and thus not even the ribosomal genes reflect true historical relationships.

Yet a few evolutionary biologists question whether the gene swapping has been as widespread as some researchers are now suggesting. TIGR's Owen White points out that the conclusion that Deinococcus has picked up plant genes is based primarily on researchers' inability to find those genes in any other microbes. But he cautions that the genome sequences currently available represent “a mere sliver” of all microbial genomes so that the genes may actually exist in a much broader range of organisms than researchers now realize. If so, the “plant” genes in Deinococcus might just be relics of genes lost from most other microbes. That might also hold for other apparent transfers.

And Peer Bork, a biochemist at the European Molecular Biology Laboratory (EMBL) in Heidelberg, Germany, argues that even if there has been a lot of lateral gene transfer, it might still be possible to detect the outlines of the tree of life. In a study published in the January issue of Nature Genetics, Bork and EMBL's Berend Snel and Martijn Huynen looked for the percentage of genes in common among 13 fully sequenced genomes, including those of bacteria, Archaea, and the eukaryote yeast. They found that each microbe shares between 77% and 17% of its genes with another organism. They didn't try to estimate what percentage of those genes had transferred in from elsewhere.

But based on those percentages, the researchers concluded that gene swapping would not displace any organism from its designated place in the existing RNA tree of life, with the kingdoms of the Archaea and eukaryotes branching off from a common ancestor, much as is now envisioned. That's because close relatives on that tree, such as Escherichia coli and Haemophilus influenzae, have higher percentages of genes in common than do organisms further apart, a finding that supports the microbes' current positions. “Gene transfer happens quite frequently at all levels,” says Bork. “But there is still a strong [evolutionary] signal in the gene content.”

## New trees needed?

Despite supportive data based on the relative gene contents in the genomes, some researchers are not convinced that the tree of life based on ribosomal genes is correct. Other types of gene comparisons and analyses of cell morphology make other evolutionary scenarios seem just as—if not more —likely, they suggest. Radhey Gupta, a molecular biologist at McMaster University in Hamilton, Ontario, for example, focuses on two different microbial features.

One feature involves small insertions or deletions, which he calls “indels,” in the coding regions of genes. When an indel exists between two conserved stretches of coding sequence, Gupta says, it almost certainly has been retained through evolutionary time, as it's highly unlikely that two unrelated organisms would have the exact same alteration. Such indels can therefore be used to trace gene lineages. Gupta also takes into account the structure of the cell membrane. The result is a drastic reorganization of the classic tree in that Gupta finds a fundamental split among the bacteria. In addition, he divides the Archaea between the two bacterial branches instead of placing them in a separate kingdom.

Gupta begins with the supposition that the most primitive organism resembled today's gram-positive bacteria (which are so called because they are dyed by the Gram stain). Based on the indels present or absent in other genes coding for highly conserved proteins, such as those involved in DNA synthesis, Gupta then came up with a scenario for how other organisms arose.

In his view, one group of these gram-positive bacteria eventually gave rise to the salt-tolerant Archaea and to the modern gram-positive bacteria, while some descendants of a second group eventually split into two lines, one of which became the gram-negative bacteria, while the other includes the Archaea that today thrive in methane- or acid-rich environments.

This picture is supported, Gupta says, by the presence in all the gram-negative bacteria examined of a particular 26-amino acid insert that is not found in the gram-positive organisms or in the Archaea. The insertion must have occurred, he says, after the gram-negative bacteria diverged from their gram-positive ancestors. Other support comes from the proportions of the bases guanine and cytosine in the genomes of these microbes, he notes. The salt-loving Archaea and many of their close gram-positive cousins belong on one major branch because they have a high proportion of these bases, while the other Archaea and gram-positive bacteria, with a low proportion, belong on the other.

In keeping with other recent evidence (Science, 13 March 1998, p. 1633), Gupta then suggests that eukaryotes arose when an archaeon fused with a gram-negative bacterium. The archaeon became the nucleus, while the gram-negative bacterium contributed most of the rest of the cell's components as indicated by the bacterial nature of the modern eukaryotic cell's membranes.

His theory got a lukewarm reception at the meeting, however. “It's an interesting hypothesis,” says TIGR's Fraser. “But I don't think the story is anywhere near that simple.” Nor does mathematician Hervé Philippe, who with microbiologist Patrick Forterre of the University of Paris-Sud has tentatively put forth an even more radical scenario. These researchers propose that a eukaryote-like cell, rather than a prokaryote, is the last common ancestor of all life-forms. “That's a more fruitful working hypothesis,” Philippe maintains.

According to the analysis Philippe presented at the meeting, one problem with previous tree-building efforts is that researchers didn't compensate sufficiently for the fact that genes can differ in their rates of evolution. (The results are also in press in Molecular Biology and Evolution and in the Journal of Molecular Evolution.) As a result, organisms that have a lot of the faster evolving genes tended to appear more ancient than they really are, confusing the trees. And Philippe says bacterial genes are the faster evolving ones, causing bacteria to fall to the tree's base.

But that's actually where eukaryotes belong, he says, basing this conclusion on an examination of the genes involved in the synthesis and use of the genetic information in the various types of organisms. He notes, for example, that many researchers think that the earliest organisms lived in an RNA world, where RNA, rather than DNA, was used to carry life's instructions. And compared to prokaryotes, eukaryotic cells have many more genes involved in processing the RNAs that may be relics of that RNA world.

In eukaryotes and Archaea, he says, the complex interactions between the proteins encoded by the genes needed to process the genetic information would have limited the ability of the genes to change. But those constraints would be less severe in bacteria, which somehow lost a few of these genes. The bacteria could have then evolved faster—and as a result they end up at the bottom of the tree when they don't deserve to be there.

This new tree of life pleases some researchers, particularly those who already have reason to suspect that the current tree might be misleading. For example, Maryland's Robb says he's often suspected that the placement of the hyperthermophiles, microbes that live at extremely high temperatures, toward the bottom of the tree might be an artifact resulting from assumptions about how fast these microbes evolve and on their initial discovery in extreme environments reminiscent of what early life might have experienced.

That placement was based in part on the finding that the DNA in the ribosomal genes of the organisms has lots of guanine and cytosine bases, an indication that their genomes had been around long enough for certain bases to become overrepresented. But it turns out that these bases tend to dominate in the genomes of organisms living in extreme environments, as they help stabilize DNA (Science, 8 January, pp. 155, 220). Thus, these DNAs do not necessarily tell the true story of evolutionary relatedness. “A lot of what Philippe is saying makes sense to me,” Robb notes.

Nonetheless, like Gupta, Philippe has his skeptics, as many researchers think it's too early to reach any firm conclusions about early evolution from the new genomes. “We just don't have that high a resolution for ancient [events],” says EMBL's Bork. To try to resolve the issue, both he and Philippe advocate not only looking at more genomes but also working to understand better how genomes change through time.

Fraser agrees that more needs to be learned about the molecular mechanisms that underlie evolution. At this point, she says, the only thing that can be said with any certainty “is that the story is by far more complicated than is suggested by the ribosomal DNA tree.”

19. MICROBES, IMMUNITY, AND DISEASE

# Borrowing--Genes--From Microbial Neighbors

1. Elizabeth Pennisi

Microbiologists have known for some time that certain pathogenic bacteria and some soil microbes swap genes, which shuttle from microbe to microbe as part of mobile bits of DNA called plasmids. Such gene transfers account, for example, for the rapid spread of resistance to antibiotics. But recent work suggests that the transfers may be even more common than thought, particularly in the deep evolutionary past, allowing microbes to exchange a wide variety of different genes—and complicating biologists' efforts to trace microbial evolution (see main text). Indeed, as new work presented at the Microbial Genomes III meeting shows, these widespread gene transfers have not been confined to the evolutionary past but are going on today, even in supposedly ancient organisms, called Archaea, whose genomes have been thought to be long fixed.

The work comes from Frank Robb and Jocelyne DiRuggiero of the University of Maryland Center of Marine Biotechnology in Baltimore and their colleagues. For the past 3 years, they and Robert Weiss's group at the University of Utah, Salt Lake City, have been sequencing the genome of Pyrococcus furiosus, a member of the Archaea that thrives in the 100°C water off Italy's Volcano Island. The researchers wanted to see how the genome of this species would stack up against those of two other pyrococci, P. horikoshii, isolated from hot springs on the ocean bottom in Japan, and P. abyssi, which came from volcanic vents in the waters off Fiji. But along the way, they found an unexpected link to a far more distant relative.

The first hint of the connection came when DiRuggiero and Winfried Boos of the University of Konstanz in Germany found that P. furiosus has the same gene for transporting the sugar maltose as does an archaeon called Thermococcus litoralis, which happens to live close by in slightly cooler waters, 85°C rather than 100. The Pacific cousins of P. furiosus have no such gene, and at first, Robb recalls, Boos and DiRuggiero “thought they had made some terrible mistake,” such as contaminating their P. furiosus DNA with T. litoralis DNA.

Further work confirmed that the link was real—and even stronger than it had seemed at first. The maltose-transporter gene in P. furiosus contains a stretch of 16,000 bases that is almost identical to a sequence in the T. litoralis gene. The researchers found identical runs of some 6000 bases in some regions of the shared section, and overall there were just 138 base differences between the two species. Because the sequences are so similar, Robb suspects the entire sequence was somehow transferred from one to the other, possibly with the help of transposons—mobile DNA elements—that flank the DNA in Pyrococcus, and that the transfer occurred within the past 1000 years, perhaps even within the past 100 years.

Researchers have come to expect aggressive gene acquisition in pathogenic bacteria that are trying to stay virulent. “But Archaea are supposed to be ancient, slowly evolving,” Robb points out. The finding that they traffic in genes today “should consolidate the idea that [early lateral transfers] could have actually happened.”