News this Week

Science  25 Jul 2003:
Vol. 301, Issue 5632, pp. 444

You are currently viewing the .

View Full Text

Log in to view the full text

Log in through your institution

Log in through your institution


    House Panel Backs 6% NSF Increase As Other Agencies Struggle

    1. Jeffrey Mervis*
    1. With reporting by Andrew Lawler, Charles Seife, and Erik Stokstad.

    In an otherwise tight budget year for research agencies, a House spending panel last week smiled on the National Science Foundation (NSF). The panel approved a 6.2% boost for NSF, the government's chief funder of nonmedical academic research, while bestowing increases of just 1% on NASA and 2.9% on the Environmental Protection Agency's (EPA's) research programs. Also last week, legislators attacked President George W. Bush's plans to research a new generation of nuclear weapons and restored proposed cuts to the U.S. Geological Survey (USGS).

    The yearlong process of passing the federal budget traditionally picks up steam in July, and this year is no exception. The House hoped to finish work this week on all but one of 13 bills for 2004, covering $785 billion in discretionary spending. [It approved a 2.5% increase for the National Institutes of Health (NIH) in June (Science, 27 June, p. 2019).] The Senate is not as far along. Congress will begin to resolve any differences between the two bodies after it returns from an August recess, giving science lobbyists another bite at the budget apple. NIH supporters, for example, hope that the Senate will endorse an 8.4% hike for the agency and that the higher level will prevail in conference.

    The $329 million increase for NSF is nearly twice the amount the president had requested. But the total, $5.64 billion, still falls well short of the $6.4 billion that advocates had sought as the first installment toward doubling the agency's budget over 5 years. Nevertheless, “it's a very positive sign” of support for basic research, says one NSF official.

    The House panel gives the green light to three new major research projects. The first is $12 million to start building two prototypes for a $391 million National Ecological Observatory Network (NEON), a proposed chain of 17 field stations. Legislators had shot down the request twice before (Science, 20 June, p. 1869), but advocates made what one congressional staffer called “a much more compelling case” this year. Even so, lawmakers told NSF not to expect any more money until it could assure them that NEON wouldn't duplicate existing federal facilities.

    Detecting movement.

    Physicists are thrilled that Congress wants to speed up the timetable for a new experiment to understand the origins of matter.


    The spending panel also approved $25 million to begin work a year early on NSF's next-generation ocean drilling program (Science, 18 April, p. 410). The money will allow NSF to start acquiring and outfitting a second drill ship, which is expected to cost $100 million. Likewise, the committee ordered NSF to spend “no less than” $8 million to start design work on a $218 million high-energy physics experiment at Brookhaven National Laboratory in Upton, New York, called Rare Symmetry Violating Processes (Science, 14 September 2001, p. 1972). “This is the best news I've heard in months,” says physicist Michael Marx, project manager for one of the instrument's two detectors.

    NSF's research account would rise by $250 million, to $4.3 billion, but its education programs would remain nearly flat, at $911 million. Legislators nixed an $8.5 million first installment on a workforce initiative and gave the foundation only 70% of the $200 million it had requested for a competitive grants program that teams universities with local schools to improve math and science instruction (Science, 11 January 2002, p. 265). NSF's program complements a similarly sized, expanding program at the Department of Education that doles out money to each state.

    NASA would receive $15.54 billion under the House proposal. Although that's a mere $150 million more than this year's budget, it's double the increase the president requested, and it includes funds for a space science mission to Jupiter. Funding levels for the space shuttle, space station, and the proposed orbital space plane are expected to be revised following the release next month of a report on the Columbia disaster from a committee led by retired U.S. Navy Adm. Howard Gehman.

    Both the House and Senate have upped the president's proposed 1% increase— $45 million—for the Department of Energy's $3.3 billion Office of Science. The House calls for a $219 million hike, whereas the Senate would add only $88 million. Much of the House increase would go to biological and environmental research and computing programs. The House also cut to $5 million the president's request for $21 million to study earth-penetrating and low-yield nuclear weapons (Science, 4 July, p. 32), while the Senate endorsed the president's request.

    For the third straight year, Congress has signaled that it will cancel proposed steep cuts to the $569 million research budget of USGS. Both houses have approved single-digit increases. Similarly, a House panel last week came to the rescue of EPA research, approving a 2.9% increase, to $767 million, instead of the 2% drop in the president's budget.

    Given the recent announcement of a projected $455 billion deficit this year, observers don't expect a spending spree when Congress returns after Labor Day to finish the 2004 budget, which officially begins on 1 October. Even so, science lobbyists are prepared to argue that research is one of the cheapest—and most effective—ways to generate the economic wealth needed to put the nation back in the black.


    Congress Wants the Twain to Meet

    1. Jennifer Couzin

    Everybody talks about the need for more research at the intersection of the physical and biological sciences. Now Congress wants the government to do something about it.

    This month, two spending panels asked the National Institutes of Health (NIH) to convene a meeting this year of every government agency that funds work in these fields. Legislators took up the cause after several scientific societies argued that more needs to be done to support work in other disciplines that have fueled biomedical breakthroughs. Favoring biology research at the expense of the physical sciences, says Mary Barkley, a past president of the Biophysical Society, means that “eventually, you'll run out of gas.”

    Dubbing their coalition “Bridging the Sciences,” the groups have enlisted the help of John Porter, a recently retired congressman and longtime friend of NIH who is now at the Washington, D.C., law firm of Hogan & Hartson. “We're just hopeful that somebody will take the lead,” says Porter, who's helped arrange meetings with NIH Director Elias Zerhouni, National Science Foundation Director Rita Colwell, and other science policy heavyweights. Adds one Hill aide, “Everybody agrees that interdisciplinary research is important. But each agency wants somebody else to do it.”

    NIH is already discussing the issue with other agencies, says James Cassatt, director of the division of cell biology and biophysics at NIH's National Institute of General Medical Sciences. One problem, he says, is that physicists are unfamiliar with the NIH grants system and, therefore, don't usually submit proposals. “I'm not sure there are limitations based on funds,” he says.

    Barkley, a biophysical chemist at Case Western Reserve University in Cleveland, agrees that physicists balk at applying for NIH support. She and Kenneth Dill, another former president of the Biophysical Society, suggest that one solution could be a new, multiagency approach to back research at the crossroads of the two disciplines.


    British Expert Leaves Impressive Arms Control Legacy

    1. Richard Stone

    CAMBRIDGE, U.K.—Earlier this week a senior judge was appointed to lead an investigation into the reported suicide of biological weapons expert David Kelly, a veteran of numerous inspections of former Soviet bioweapons facilities and of 37 inspections in Iraq during the 1990s. Kelly had become embroiled in an ugly spat between the government and the BBC over controversial news reports that officials had “sexed up” intelligence reports on Iraq prior to the war. Kelly's death closely followed the revelation last week that he was the principal source of the BBC reports. The tragedy has left Kelly's colleagues not only saddened but perplexed that someone who proved so quietly determined in dealings with evasive officials in Russia and Iraq could have become so boxed in.

    Kelly, a microbiologist by training and a senior adviser to the Proliferation and Arms Control Secretariat of the U.K.'s Ministry of Defence, was widely respected for his expertise and his courteous, but forceful, dealings with adversaries bent on hiding illicit bioweapons activities. As one of the chief weapons inspectors in Iraq, Kelly made one of the biggest discoveries of his life. In the early 1990s, searching for evidence of an offensive bioweapons effort in Iraq, Kelly and U.S. colleague Richard Spertzel noticed something suspicious: A few years earlier, Iraq had gone on a buying spree, importing 39 tons of bacterial growth media. Officials produced documents claiming that the agar was for hospitals to diagnose infections. But when the inspectors compared Iraqi imports with those into neighboring countries Iran and Syria, figuring they should be similar, “it was clear that Iraq's imports were way too high,” Kelly said in an interview with Science shortly before his death. In addition, the agar's bulk packaging did not correspond with its intended use. The inspectors accused Iraqi officials of forging the documents and importing the agar for the production of anthrax and other strains, forcing them in 1995 to acknowledge for the first time that Iraq had pursued a clandestine offensive bioweapons program.

    Important preparation for his later roles came during his tenure, from 1984 to 1992, as head of microbiology at the Chemical and Biological Defence Establishment in Porton Down. When Kelly arrived at the former weapons laboratory, it had only a skeleton crew of microbiologists, and they were primarily involved in the decontamination of Scotland's Gruinard Island, where Great Britain had conducted tests with weaponized anthrax during World War II.

    Kelly made a scientific case for doubling the division's resources to step up biodefense R&D and in 1986 was granted 2 years' funding to demonstrate that it would work. Kelly recruited several young scientists and, “through his enthusiasm, energized the team,” says Graham Pearson, head of Porton Down at the time. The division got high marks in the 2-year review, and by the 1991 Gulf War, Pearson says, “we were in a position to deploy a limited biodefense capability.” It's thanks to Kelly, Pearson says, that “Porton Down today has world-class facilities.”

    Word of Kelly's achievements reached Rolf Ekeus, the first head of the United Nations Special Commission (UNSCOM) to investigate Iraq's illicit weapons programs. He tapped Kelly to lead the first bioweapons team in Iraq in August 1991. Pearson, who wrote an authoritative account of the UNSCOM years, says that Kelly's knowledge, coupled with “his persistent yet polite questioning” of Iraqi personnel, “helped to uncover much of what was being hidden by Iraq.”

    Kelly also was a key player in efforts in the early 1990s to ferret out the extent of the Soviet Union's offensive bioweapons efforts. After key details of the program emerged from two defectors in the dying days of the Soviet Union, the United Kingdom, the United States, and a grudging Russia signed a trilateral agreement in 1992 that called for inspections at facilities suspected of being engaged in recent bioweapons activities. The initiative unraveled in the mid-1990s due to Russia's reluctance to come clean on its past activities and refusal to permit inspections of military labs. Kelly, the only expert to have taken part in all the trilateral site visits, had warned recently that Russia has yet to demonstrate convincingly that it has abandoned its offensive bioweapons program.

    It's unclear whether the forthcoming inquiry will provide closure for Kelly's colleagues. “He was one of the finest and most dedicated men I have known,” says former UNSCOM inspector Debra Krikorian, who also worked with Kelly on the trilateral initiative. Adds Pearson, “He will be sorely missed, as his knowledge and expertise were truly unique.”


    U.K. Government Panel Gives GM Crops Cautious Support

    1. John Pickrell*
    1. John Pickrell is a science reporter based in London, U.K.

    LONDON, U.K.—A panel of experts appointed by the British government to assess the safety of genetically modified crops gave a qualified thumbs-up earlier this week in a report* that may pave the way for a breach in the 5-year Europe-wide moratorium on commercial GM crop planting. In the most comprehensive scientific review yet attempted—filling 300 pages and with roughly 600 references to scientific studies—the 24 scientists and other experts argued that there is no scientific basis for ruling out GM crops but that there is still a need for tight regulation and much further research to fill knowledge gaps. “This is not a red light or a green light for GM crops,” says chemist David King, the government's chief scientific adviser, who chaired the panel.

    The United Kingdom, along with the rest of the European Union, has had a moratorium on commercial GM crops since 1998. In spring 2002, in the face of largely hostile public opinion, the British government commissioned “three strands of evidence”— including this review—and it hopes to make a final decision later this year on whether to give GM crops the go-ahead. The other strands are a series of public debates (Science, 13 June, p. 1637) and an economic report released last month. That report argued that with few GM crops currently suited to British conditions, short-term financial benefits would be minimal but long-term profits could be large.

    The new report noted that there has been little evidence of harm to human health in countries where GM crops have been grown over the past 7 years, nor is it likely that these crops will invade the countryside. But the panel concluded that the most important issue is the potentially detrimental effect on farmland biodiversity and wildlife, and it called for more research on biodiversity, soil ecology, allergenicity, and gene flow to non-GM plants, among other issues. Each crop should be considered on a case-by-case basis, says King. The experts postponed a final decision on these issues until results from government-funded farm trials are published later this year. These studies—the largest ever attempted—are testing the effect of four different GM crops on wild plant and insect diversity.

    No green light.

    U.K. government science adviser David King.


    To forestall criticism of the report, the panel members came from a wide range of backgrounds, including representatives of biotech giants Monsanto and Syngenta alongside experts from a conservation organization, the Royal Society for the Protection of Birds, and from the Green Alliance pressure group. Biologists, sociologists, and ethicists from numerous universities also participated.

    Sue Mayer of the genetic technology watchdog GeneWatch UK commended the review for bringing “the issue of uncertainty to the fore.” Uncertainties include ways to detect any potential harm to human health and the lack of accurate predictions of environmental impacts. However, “the biggest concern,” in her view, is that the report plays down some issues, such as differences between genetic modification and traditional breeding. Others disagree. “The risks associated with GM in terms of food safety and the creation of so-called superweeds are minimal,” says Jim Smith, a GM food expert and spokesperson for London's Royal Society, although he acknowledges that a number of questions remain about how GM technology might affect wild plants and animals.


    Depth Charges Aimed at Columbia's 'Submarine Patent'

    1. Eliot Marshall

    In 1983, Columbia University secured a fundamental biotechnology patent that has netted hundreds of millions of dollars. The patent has been such a money spinner that the university couldn't bear to see it expire, so after trying unsuccessfully to get Congress to extend its life, Columbia persuaded the U.S. Patent and Trademark Office to issue a second patent last year on the same invention. That, at least, is how five companies see it. They have blitzed the university with simultaneous lawsuits in three federal courts. University officials reject the allegations, saying they are just trying to collect on a valid new patent.

    The legal jousting centers on a technique for inserting genes into a cell, invented in 1980 by Columbia biologist Richard Axel with colleagues Michael Wigler and Saul Silverstein. Based on their research, the university won a U.S. patent (number 4,399,216) for a “cotransformation” process that inserts two pieces of DNA, one of which expresses the desired protein and the other a “marker” protein that can be used to identify which cells have taken up the DNA. The technique, which makes it simpler to select the transformed cells before growing them in quantity, has become a mainstay of research and commercial production.

    The original patent expired in 2000. While it was in force, Genentech of South San Francisco estimates that it paid Columbia royalties of more than $70 million. Biogen of Cambridge, Massachusetts, estimates it paid $35 million; Genzyme, also of Cambridge, paid almost $25 million. The total amount paid by about 30 companies, according to a complaint by Biogen and others, is in the “hundreds of millions of dollars.”

    Renewable resource.

    Richard Axel's 1980 research netted Columbia University similar patents in 1983 and 2002.


    Last fall, Columbia obtained a new, 17-year U.S. patent (6,455,275) using claims based on Axel's original 1980 research. When Columbia began asking biotech companies for new payments this year, they rebelled. Biogen joined with two other firms in Massachusetts—Genzyme and the Abbott Bioresearch Center—to sue the university last week for what they describe as “an illegitimate effort to extend its patent monopoly” on the Axel invention. Filed in federal court in Boston on 15 July, this complaint asks that the patent be declared invalid because it was issued in error. The complaint also seeks legal costs. It echoes complaints filed by Genentech and Amgen of Thousand Oaks, California, on 15 April in San Francisco and on 18 June in Los Angeles, respectively.


    Genzyme's legal counsel, Thomas DesRosier, argues that the university “pulled a bait and switch” on the patent office, using legal gambits to keep alive a patent on a discovery that is “substantially the same” as the one whose patent died in 2000. The Boston complaint calls the result a “submarine patent,” because it lay dormant for years but burst on a mature industry last fall with new royalty demands. It states that Columbia used “dilatory tactics” to win the new patent and charges that it previously waged a “widely criticized campaign” in Congress to boost the life of the Axel discovery. Senator Judd Gregg (R-NH), a Columbia alumnus, tried to amend a bill in 2000 to extend the patent's life by 15 months, but the proposal died amid controversy.

    In response to Genentech's suit, the university wrote that the new patent is “valid and enforceable.” University senior executive vice president Robert Kasdin declines to discuss the case in detail, saying only that “the U.S. Patent and Trademark Office reviewed a patent application filed by us in 1995 and concluded that it reflected a new invention. … We believe that their conclusion is sound.” The biotech firms are waging legal warfare, an official says, because they “just don't want to pay” the royalties they owe.

    The differences between key parts of Columbia's 1983 and 2002 patents may appear at first glance to be “more semantic than substantive,” says independent patent attorney Richard Osman of Hillsborough, California. But he warns, “It is imperative to consider the significance of every word of each claim … before jumping to any conclusions.” It won't be an easy case for the plaintiffs to make, says Osman: They “must convince the judge or jury that the patent office made a mistake.”

    Like others trying to stake out intellectual property rights in the fiercely competitive biotech world, Columbia is learning that an aggressive strategy can bring in lucrative patents—and some blockbuster legal bills.


    Dark Energy Passes Another Test

    1. Charles Seife

    Dark energy, the baffling antigravity force that is flinging the universe apart, is here to stay. Scientists from nearly two dozen institutions across the globe have announced what is arguably the most direct detection of dark energy to date, leaving little hope for those who prefer the simplicity of a universe without the enigmatic energy.

    “It's exciting because it's more or less a direct detection of dark energy,” says Wayne Hu, a cosmologist at the University of Chicago.

    The researchers spotted evidence of the “integrated Sachs-Wolfe” (ISW) effect, a quirk of general relativity that causes massive objects such as galaxy clusters to change the temperature of light passing nearby. Einstein's theory says that a large chunk of matter creates a dimple in the fabric of spacetime. If a photon falls into that dimple, it gains energy just like a ball rolling down a hill. As the photon climbs out of the pit, it loses energy again. If the incline is the same size as the decline, then the photon loses as much energy as it gains and winds up the same as when it came in.

    However, dark energy puts a wrench in the works. It causes the fabric of spacetime (and the dimple itself) to stretch as time passes, so the photon climbs out of a pit that is slightly shallower than the one it fell into many years earlier. The photon therefore gains more energy falling in than it loses climbing out. At the same time, the flattening of the spacetime dimple squeezes the light inside it, raising its frequency and thus its temperature (see figure). Both effects result in a photon that is more energetic and has a higher temperature after passing near a massive object—the ISW effect. The equations of general relativity show that the effect can occur only if dark energy is a big fraction of all the stuff in the cosmos.

    Hot plate.

    As spacetime hollows flatten, light crossing them increases in frequency and energy.


    To spot the ISW effect, scientists need to compare light that passed by large chunks of mass with light that steered well clear of gravity wells. This job became a lot easier this spring, when the Wilkinson Microwave Anisotropy Probe (WMAP) satellite made an incredibly detailed map of the cosmic microwave background—ubiquitous cool light from the ancient universe (Science, 2 May, p. 730). Now, in a paper submitted to Physical Review Letters, scientists have taken that map and compared it with a growing survey of galaxies known as the Sloan Digital Sky Survey (SDSS). The paper shows that, indeed, the photons from the cosmic microwave background that passed close to galaxies (and thus fell into and out of gravitational wells) were hotter, on average, than those that were far away from galaxies and dimples in spacetime—clear evidence for the ISW effect and dark energy.

    “It's really the marriage of two major new surveys that allows you to do this,” says team member Robert Nichol, a physicist at Carnegie Mellon University in Pittsburgh. The statistical confidence in the discovery is somewhere between two and three sigma—about 99%, which is fairly low by physics standards. Nevertheless, Nichol says, internal checks show that the analysis is “quite robust.” Add to that several other papers that compared WMAP data to older surveys of radio and visible sources in the sky, and scientists are confident that the ISW effect has been detected.

    By the time the SDSS project finishes its work in 2006, it should have about two and a half times as much data as it does now—enough for team members not only to detect the ISW effect with higher confidence but also to measure how much dark energy there is and throw some light on dark energy's nature. “ISW is very sensitive to the amount of dark energy, so you can make a precise measurement of that,” says Nichol. “And you will be able to measure not only the amount but the equation of state,” a key parameter that captures many of the properties of dark energy, such as whether it changes strength over time. With luck, the weird stuff that makes up three-fourths of the universe may even start to feel familiar.


    Late Date for Siberian Site Challenges Bering Pathway

    1. Richard Stone

    As elusive as the Cheshire Cat, the first people to arrive in the Americas have tended to appear and vanish with each new twist in the archaeological record. The latest disappearing act may be taking place on page 501, where new evidence, some claim, casts another shadow over a once-cherished idea: that Asian big-game hunters crossed the Bering Land Bridge to give rise to the Clovis people, who were considered the first Americans. New dates show that a crucial Siberian site, thought to be a way station along the Bering road, wasn't occupied until after the Clovis had begun killing mammoths in North America.

    The new finding “removes what was, until now, the critical link in the chain connecting Clovis to Siberia,” says David Meltzer, an archaeologist at Southern Methodist University in Dallas. The paper “is really thought-provoking,” adds Richard Potts, director of the human origins program at the Smithsonian Institution's National Museum of Natural History (NMNH) in Washington, D.C. “Maybe it's time to consider less obvious sources for the oldest migration to the continent,” he says.

    Who the earliest Americans were and how they got there is one of anthropology's biggest riddles. The first universally acknowledged culture in the Americas is that of the Clovis, who scattered their distinctively fluted projectile tips across North America starting about 13,600 years ago (using corrected radiocarbon dates) before vanishing several centuries later. Their ancestors were thought to have crossed the land bridge that sometimes linked present-day Alaska and northeastern Siberia; the bridge appeared and disappeared with the fall and rise of sea level during the last Ice Age (see map). But a handful of pre-Clovis sites, including one in Monte Verde, Chile, dated to about 15,000 years ago, have challenged this idea (Science, 2 March 2001, p. 1730).

    Land grab.

    Lower sea level during the Ice Age exposed more land in Beringia.

    For researchers who believe that the Clovis were truly the first Americans, the site of Ushki Lake on Russia's Kamchatka Peninsula has been a critical piece of evidence. One of the few late Ice Age sites in northeastern Asia, it was discovered in 1964 by Siberian archaeologist Nikolai Dikov, who ran a secretive dig for 25 years. “He rarely invited anyone out to the site apart from trusted colleagues and students,” says paleoanthropologist Ted Goebel of the University of Nevada, Reno. But word began to filter out, and in 1978 Dikov, in his first English publication on Ushki, offered some brief but tantalizing descriptions about what he'd found. One layer at the site, dated to 12,600 years ago, included wedge-shaped cores, tiny stone blades, and burins: pointed tools for carving bone and antler. And beneath the floor of an earthen shelter, next to human bones, Dikov also discovered a very different collection of implements, including flaked cores and chipped bifacial points, and stone beads that he called wampum. Most striking was the date: Charcoal in the grave was dated to 16,800 years ago. If Dikov was right, Ushki's earliest inhabitants—even though their stone points are shaped differently from those of the Clovis—might have provided one ancestral Clovis strand.

    Wondering if the site had more secrets to reveal, Goebel contacted Dikov's widow, archaeologist Margarita Dikova of the Northeast Asian Interdisciplinary Research Science Center in Magadan, Russia, who agreed to a joint expedition in 2000. Joined by Michael Waters of Texas A&M University in College Station, they spent 3 weeks sampling and mapping the sediments at Ushki Lake. Goebel came away with a new appreciation of Dikov's work. “Everything he did was impeccable,” he says. Everything except one: New analyses of charcoal fragments showed that the grave is only about 13,000 years old—6 centuries later than the first Clovis points. “I was pretty shocked,” says Goebel.

    Thus the Ushki Lake inhabitants themselves cannot be ancestral to the Clovis people. But to Goebel and his team, Ushki's dating facelift doesn't rule out Clovis origins in Beringia. Goebel notes that the most ancient known Beringians are now those of the Nenana culture, who fashioned small biface points and knives; their oldest site, Broken Mammoth in central Alaska, is dated to 14,000 years ago, although no one is certain when or from where they reached Alaska.

    Uncovering Ushki's secrets.

    The site of Ushki Lake lies in the shadow of Kliuchevskoi Volcano. Inset: Team member Igor Sleptsov excavates a 12,000-year-old hearth.


    It's possible, says Goebel, that these Beringians raced down into the North American plains in a few centuries, developing into Clovis as they went—a sprint many consider worthy of the Iditarod. “I do not see this as a dilemma,” says eminent geochronologist C. Vance Haynes of the University of Arizona in Tucson. He notes that recent dating suggests that the Clovis culture may be a few centuries younger, making the run from Alaska somewhat more plausible. According to Haynes, mammoths and other large animals encountered along the way would have required new hunting strategies and weapon designs that may have spurred the development of the Clovis point.

    The other possibility, the authors say, is that there is no link between Clovis and the cultures of Ushki and Nenana, opening up a wide variety of scenarios for the peopling of the Americas. Some Bering enthusiasts favor a much earlier migration across the land bridge—before the last glacial maximum 24,000 years ago—leaving ample time to reach Monte Verde. “I think with time we will find a link between Siberia and America, but it will be a much older link,” Waters says. The newly discovered Yana Rhinoceros Horn Site in northern Siberia, claimed to be 25,000 or more years old, may support that idea. Others disagree. “There's no evidence as yet in either Siberia or the Americas to support this,” says NMNH anthropologist Dennis Stanford.

    With NMNH colleague Bruce Bradley, Stanford is a leading proponent of multiple entry points for early Americans, including the once-heretical idea of an ancient North Atlantic crossing. Stanford notes that points and blades from a controversial site at Cactus Hill, Virginia, “show remarkable correspondence to Clovis” and may be as much as 18,000 years old. The artifacts also closely resemble the Solutrean technology of northern Spain from around that time. “If artifacts resembling Solutrean were found in western Beringia, most archaeologists would propose an ancestral relationship with Clovis,” says Stanford, who argues that “we must look outside of northeast Asia” for the origins of the earliest Americans. Meltzer disagrees. Ushki's fall, he argues, “does not mean we should give up on Siberia and go looking for Clovis origins in all the wrong places”—which to him means Europe.

    One thing the experts do agree on is that the archaeological record is too sparse to settle the debate. “This isn't a problem we can think our way out of,” says Meltzer, who urges a redoubling of efforts in Siberia. “We need more early sites and data.” The link between Clovis and Ushki Lake may be fading, but for many, the Siberian connection hasn't lost its Cheshire grin.


    New Count of Old Whales Adds Up to Big Debate

    1. Naomi Lubick

    A new census of past whale populations has produced a controversial result. The study, using genetic markers, suggests that some whale populations were once about 10 times larger than previously suspected. The results hint that overhunting is the culprit and that the ocean can sustain more of the leviathans than previously thought. But some scientists question the findings, which have implications for international management practices: If the results are true, goals for whale recovery efforts should be set higher, the authors of the new work say.

    Whale populations are known to be volatile; they have risen and plummeted with changes in market demands, hunting technology, and ocean conditions. To gauge populations, researchers traditionally relied on the sometimes spotty logbooks compiled by whaling captains over the last few centuries. Such records helped the International Whaling Commission peg the past North Atlantic population of humpback whales at about 20,000, for example, which is twice the current number.

    But according to an analysis on page 508, North Atlantic humpbacks once numbered 240,000. Marine conservation biologists and geneticists Joe Roman of Harvard University and Stephen Palumbi of Stanford University used a small section of mitochondrial DNA (mtDNA) to reconstruct whale history. They compiled a database of mtDNA samples from humpback, minke, and fin whales mostly in the North Atlantic, with some data coming from the Southern Hemisphere and the Antarctic.

    The team then compared the genetic distance between humpback and fin whales, which diverged at the genus level between 6 million and 10 million years ago, to estimate how quickly mtDNA had mutated. They also measured the genetic diversity within each species of whale. Given those numbers, they could calculate how many breeding females would have been necessary to accumulate the genetic variability observed in the mtDNA samples. (mtDNA is passed down through eggs, not sperm.) A final calculation, based on the proportion of a whale population that consists of breeding females, produced an estimate of the total population.

    Whale of a dispute.

    According to a genetic analysis, many more humpback (top) and fin (bottom) whales once populated the North Atlantic than previously suspected.


    Roman and Palumbi calculate that fin whales in the North Atlantic once numbered about 360,000 individuals. That's about six times more than previously thought. The North Atlantic minke whale population was once 265,000, or about twice today's population. The results “require a fundamental rethinking of the oceans” and their capacity to support large numbers of whales, says Palumbi.

    Although the researchers admit that these populations might have only existed hundreds of thousands of years ago, they speculate that huge numbers of whales roamed the oceans until relatively recently, when industrialized hunting began taking its toll.

    But many geneticists and whale biologists object. “The proposed reduction in abundance could have happened at any point in evolutionary time,” says Per Palsbøll, a whale population geneticist at the University of California, Berkeley. That calls into question the work's relevance to present management efforts, he says, “since the abundance estimate may relate to a rather different time period with a different climate and ecological conditions.”

    Another key uncertainty is tied to setting the molecular clock of mtDNA. Karen Martien, a conservation geneticist and fisheries biologist at the Southwest Fisheries Science Center in La Jolla, California, disagrees with Roman and Palumbi's estimated mtDNA mutation rates. “Some [mtDNA] base pairs might have mutated multiple times,” she says, but the researchers would register only one mutation, which would artificially inflate population estimates.

    Palumbi acknowledges that the piece of mtDNA they used is “particularly badly behaved,” and the two researchers plan to use other genetic loci to confirm their findings, as well as data from Southern Hemisphere populations of whales. Nevertheless, Palumbi says that mutation rates would have to be five times higher and the duration of a generation at least twice as long to make the genetic analysis fit with conventional population estimates.

    But some researchers aren't ready to abandon estimates based on logbook records. “From what we know of the recent history of these populations, we simply cannot reconcile [Roman and Palumbi's] figures with anything that's even remotely realistic,” says Phillip Clapham, a whale biologist at the Northeast Fisheries Science Center in Woods Hole, Massachusetts.

    Precipitous population drops are well documented in other large mammals, such as brown bears, points out Peter Beerli of Florida State University in Tallahassee, who developed the population estimate model used by Roman and Palumbi. The team's mutation rate is reasonable for mammals, he says, but other variables are uncertain enough to have thrown off their confidence intervals. Still, he believes “the numbers are not that wrong.”


    A Boost for Tumor Starvation

    1. Jean Marx

    Despite some setbacks, researchers developing antiangiogenesis treatments for cancer have persevered. Now the strategy is racking up new successes and earning credit even for conventional treatments

    The past few years have been tough on companies trying to develop drugs that fight cancer by attacking the blood vessels that feed tumors. Although these so-called antiangiogenesis drugs produced some small benefits in early clinical trials, they didn't live up to their publicity, a development that helped pull down biotech stock prices. But the field appears on the verge of a comeback.

    Perhaps the best sign that antiangiogenesis research is finally paying off came in early June at the annual meeting of the American Society of Clinical Oncologists (ASCO). The results of a large clinical trial presented there showed that an antiangiogenesis drug called Avastin, made by Genentech Inc. of South San Francisco, prolonged the lives of patients with advanced colon cancer. Those results “put questions about the field to rest for a while,” says cancer researcher Lee Rosen of Targenix in Los Angeles.

    Angiogenesis researchers welcomed the news; Avastin had failed to extend the lives of breast cancer patients in an earlier clinical trial. Before the colon cancer findings were unveiled, Robert Kerbel, who studies angiogenesis at Sunnybrook and Women's College Health Sciences Centre in Toronto, had quipped, “If that result is negative, I'll be working in a delicatessen.” He won't be trading his lab coat for an apron: The result was “better than we thought it would be,” he says.

    Now, researchers are anticipating more successes. William Li, who heads the Angiogenesis Foundation, a nonprofit in Cambridge, Massachusetts, says that more than 60 antiangiogenesis drugs are currently in clinical trials. They include naturally occurring substances, such as the once-ballyhooed proteins endostatin and angiostatin; newly designed synthetic drugs; and older chemotherapy drugs and other medicines that have unexpectedly turned out to have antiangiogenic effects. Even standard radiation therapy apparently kills tumors partly by destroying their blood vessels (Science, 16 May, p. 1155).

    But although the antiangiogenesis field is currently looking up, researchers know that they have a long slog ahead. The drugs appear to work best in combination with other cancer therapies, and a great deal of work will be needed to find out just how effective the various drugs are and how best to use them. Development of cancer therapies tends to advance by baby steps rather than giant leaps.

    Indeed, one reason for the early disappointment may have been the impossibly high expectations raised by an article that appeared in The New York Times just over 5 years ago. Although the necessary caveats were there, it included a quote by double helix co-discoverer James Watson to the effect that endostatin and angiostatin—neither of which had been tested in humans—were going to cure cancer in 2 years (Science, 15 May 1998, p. 996). Needless to say, that hasn't happened.

    Timing is everything

    Endostatin and angiostatin were discovered by Judah Folkman's team at Children's Hospital in Boston. Many teams, including Folkman's, found that the proteins shrink mouse tumors, but doubts arose early last year when some researchers were unable to replicate some of the endostatin work (Science, 22 March 2002, p. 2198). Even so, the animal results have been sufficiently promising that endostatin has so far been tested in about 140 patients and angiostatin in another 65. Although the proteins have induced tumor regression or stabilization in some, the preliminary results suggest that the proteins won't deliver a knockout blow to cancer, at least not by themselves. Even Folkman concedes that the clinical work has “stalled a bit.”

    It may soon come to a halt. A few weeks before the Avastin colon cancer results came out, the company that produces endostatin and angiostatin, EntreMed Inc. of Rockville, Maryland, citing financial difficulties, announced that it was going to stop making the proteins. Unless another company acquires the rights to the proteins and proceeds with their development, they will not become cancer drugs.

    Avastin, an antibody that blocks a key angiogenesis instigator, has fared better, making it into the third and final stage of clinical testing. The prime goal of this stage is assessment of clinical efficacy—and Avastin failed that test in the earlier breast cancer trial before passing it in a test of colon cancer patients.

    Avastin mops up a protein called vascular endothelial growth factor (VEGF). Identified roughly 15 years ago, VEGF is a prime instigator of the formation of new blood vessels. In 1993, Genentech's Napoleone Ferrara, a VEGF co-discoverer, and colleagues showed that a monoclonal antibody to VEGF could block the growth of human tumors transplanted into mice. “That was the first time anyone could show conclusively that inhibiting angiogenesis inhibits tumor growth,” Ferrara says. Until then, the evidence that tumors require new blood vessel formation to survive was largely indirect.

    The reason the anti-VEGF antibody Avastin worked in the colon cancer trial but not in the breast cancer trial may relate to differences in the two groups of patients and in the natures of their cancers. In both trials, patients had advanced disease that had metastasized to new sites in the body. They were given standard chemotherapy either with or without Avastin. In the breast cancer trial, there were intriguing signs that Avastin shrank some of the tumors for a time, but it did not prolong the patients' lives.

    In contrast, as reported at the ASCO meeting by Herbert Hurwitz of Duke University, the 403 colon cancer patients who received Avastin had a median survival of 20.3 months compared with 15.6 months for the patients who got only conventional chemotherapy. What's more, Avastin's side effects were relatively mild for a cancer drug.

    Researchers suspect that the colon cancer patients fared better on Avastin because their tumors are more dependent on VEGF-induced angiogenesis than are those of the breast cancer patients. Tumor cells produce angiogenesis-promoting proteins, and a few years ago Adrian Harris, Roy Bicknell, and their colleagues at John Radcliffe Hospital in Oxford, U.K., found that early breast cancers produce mainly VEGF. More advanced cancers apparently up the ante, however, pouring out additional angiogenic proteins, including fibroblast growth factors (FGFs), placental growth factor, and transforming growth factor beta. The breast cancer patients' tumors thus may have been able to circumvent Avastin's blockage of VEGF by means of these other proteins.

    Patients with metastatic colon cancer don't usually survive as long as breast cancer patients do. They have fewer options for life-extending therapy, and so their tumors may have effectively been “younger” and more VEGF-dependent. Folkman and others suggest that Avastin might have fared better in the breast cancer trial if it had included only those patients whose tumors were dependent on VEGF.

    Researchers studying animal cancer models are also finding that the timing of anti-VEGF therapy can be important. One recent example comes from Gabriele Bergers, Douglas Hanahan, and their colleagues at the University of California, San Francisco (UCSF), who work with a mouse model genetically engineered to develop pancreatic tumors. The team tested SU5416, a drug from Sugen Inc. in South San Francisco, that blocks VEGF activity by inhibiting the receptor through which it exerts its effects. They found that the drug blocks early development of pancreatic tumors but has little effect on late-stage tumors.

    Another Sugen drug, SU6668, had just the opposite effect. The UCSF researchers found that it had little ability to stop the growth of early tumors but caused marked shrinkage of late-stage tumors. This suggests, Bergers says, that “the vasculature in large [tumors] is different from that in small ones.”

    Further results, in the May issue of the Journal of Clinical Investigation, point to a possible explanation. SU6668 has broader powers than SU5416, blocking the receptor for another angiogenesis-promoting protein, platelet-derived growth factor (PDGF), as well as the VEGF receptor. In the mouse tumors, PDGF receptors are found not on the endothelial cells that form the blood vessel walls but on accessory cells called pericytes that apparently help maintain vessel integrity. This suggests that the pericytes are the cells disrupted by SU6668 in older tumors, which Bergers and her colleagues confirmed microscopically.

    Intriguingly, the UCSF team found that treating mice with both SU5416 and SU6668 reduced tumor burden more than did either drug alone. So did combining SU5416 with Gleevec, a drug already approved for cancer therapy whose cellular targets include the PDGF receptor. These observations tie in with another lesson emerging from the past 10 years of testing antiangiogenesis therapies: To get the best results, it may be necessary to combine antiangiogenesis drugs, either with each other or with conventional chemotherapy drugs. “Angiogenesis is such a diverse process, no one agent will work by itself,” says oncologist Roy Herbst of the University of Texas M. D. Anderson Cancer Center in Houston.

    The antiangiogenesis pantry

    As things are shaping up, clinicians will have plenty of antiangiogenesis drugs to choose from. Their mechanisms of action are equally diverse. VEGF is a favorite target. In addition to Avastin and the Sugen drugs, drugs directed at blocking VEGF activity include a protein called VEGF-Trap, designed by scientists at Regeneron Pharmaceuticals in Tarrytown, New York. It is a hybrid protein, containing segments of two different VEGF receptors, that binds extremely tightly to VEGF.

    Angiogenesis in action.

    In this micrograph, blood vessels grow toward a sarcoma tumor (dark area at right) in rat muscle. This contrasts with the normal gridlike pattern of blood vessels that appears at upper left.


    An indication of VEGF-Trap's effectiveness comes from Jessica Kandel and Darrell Yamashiro of Columbia University College of Physicians and Surgeons in New York City and their colleagues. In previous work, they found that human neuroblastoma tumors transplanted into mouse kidneys respond poorly to Avastin. New results, published online on 24 June in the Proceedings of the National Academy of Sciences (PNAS), show why: When the tumor transplants begin to grow, they take over existing kidney blood vessels. In the absence of treatment, the growing neuroblastomas eventually induce formation of new blood vessels and eliminate the co-opted ones.

    Although Avastin can prevent this new blood vessel growth, it apparently has little effect on the original vessels: They persist and allow the tumors to survive the treatment. In contrast, treatment with VEGF-Trap eliminates all the blood vessels, both co-opted and new. Kandel and her colleagues speculate that it sops up VEGF so completely that it eliminates even the small quantities needed for vessel maintenance. Early clinical trials are under way to see whether VEGF-Trap can be given safely to human patients.

    But VEGF is only one of several angiogenesis-promoting molecules in researchers' sights. Drugs have also been designed to inhibit enzymes called matrix metalloproteinases that help endothelial cells migrate to the sites of new blood vessel formation. And still others are aimed at blocking the activities of certain integrins. When these proteins, which are located on endothelial cell surfaces, detect angiogenic stimuli, they transmit signals that tell the cells to proliferate and migrate for vessel formation. Raghu Kalluri and his colleagues at Beth Israel Deaconess Medical Center in Boston reported in the 15 April issue of PNAS that endostatin and another naturally occurring antiangiogenic protein, tumstatin, work partly by blocking integrin signaling.

    It's also become clear over the past few years that many drugs produced without any thought of blocking angiogenesis—or even treating cancer for that matter—have antiangiogenic effects. These include the anti-inflammatory drugs Celecoxib and Rofecoxib as well as thalidomide, a drug once taken off the market because it causes birth defects that is now used to treat multiple myeloma, a cancer of the bone marrow. The antiangiogenic effects of these drugs, all of which are now in clinical cancer trials, were “big surprises,” notes Folkman.

    Both new and old-style chemotherapy drugs have provided their share of surprises, too. In the new category are some of the drugs designed to counter the actions of one or another of the numerous cancer-causing oncogenes that scientists have identified over the past decades. A few years ago, for example, Kerbel became puzzled by an anomalous action of the drug Erbitux, made by ImClone Systems in New York City. The drug and ImClone are perhaps best known for the insider-trading scandal that sent company founder Samuel Waksal to jail and ensnared domestic diva Martha Stewart, who was recently indicted.

    That sideshow notwithstanding, the science underlying Erbitux is sound. It's an antibody that blocks a receptor for epidermal growth factor (EGF). Contrary to expectations, however, researchers had found that although the drug shrinks tumors effectively in mice, it has only modest effects on tumor cells in culture. “Usually,” Kerbel says, “it's the other way around.”

    View this table:

    He hypothesized that Erbitux might exert some of its effect on tumor blood vessels—an effect that wouldn't come into play in cultured cells. Subsequent experiments showed that tumor cells secrete VEGF in response to EGF. And Kerbel says, “We found significant reduction of VEGF secretion in tumors treated with Erbitux and also [found] reduced blood-vessel density.” Since then, Kerbel's group and others have shown that several oncogenic proteins promote cancer development partly by fostering angiogenesis. They do this both by triggering the release of pro-angiogenic proteins and by inhibiting production of natural angiogenesis suppressors.

    Conventional chemotherapeutic drugs, which kill tumor cells directly, also have unexpected antiangiogenic effects. Here, too, timing is important. The usual approach to chemotherapy is to give patients the maximum dose they can tolerate and then allow them a few weeks of recovery before administering another round. But 3 years ago, Timothy Browder, a postdoc in the Folkman lab, found that that schedule can mask the drugs' antiangiogenic effects because the tumor blood vessels grow back during the interval between doses (Science, 14 April 2000, p. 245).

    Both the Folkman and Kerbel teams then found that they could prevent vessel regrowth by giving smaller doses of chemotherapy drugs more frequently. Such a regimen, sometimes called metronomic chemotherapy or chemotherapy-lite, could inhibit tumor growth in mice, although only temporarily. But the researchers found that they could prolong the antitumor effects by combining low-dose chemotherapy with an antiangiogenic drug. “You can maximize the effects of conventional chemotherapy by neutralizing factors like VEGF that might compromise [chemotherapy's] endothelial effects,” Kerbel says.

    Similarly, radiation therapy shrinks tumors at least partly by destroying their blood vessels. In previous work, cancer biologist Richard Kolesnick and radiologist Zvi Fuks of Memorial Sloan-Kettering Cancer Center in New York City found that radiation damages the gastrointestinal tract by triggering a form of cell suicide called apoptosis in the tract's blood vessels. The researchers found that this activity depends on an enzyme called acid sphingomyelinase (asmase). They reported in May that when they knocked out the asmase gene in mice, tumors transplanted into the animals “grew much faster and became radiation resistant,” says Kolesnick. The researchers linked this resistance to reduced apoptosis in the tumor blood vessels of the knockout mice.

    In addition, work on animal models by Ralph Weichselbaum of the University of Chicago, among others, has shown that radiation's antitumor effects can be enhanced by adding an antiangiogenic drug to the therapeutic regimen. “There is a lot of merit” to combining agents that attack DNA, including radiation and some chemotherapy drugs, with antiangiogenesis therapy, Weichselbaum concludes.

    After learning that a newspaper had published his obituary, a very much alive Mark Twain responded that “the report of my death was an exaggeration.” The same can be said for antiangiogenesis research. No one is predicting cancer cures within the next 2 years anymore. But as researchers figure out how to design antiangiogenesis drugs, time their administration, and combine them with other therapies, Folkman says, “it will be very exciting over the next 10 years to see how this plays out.”


    NIH Program Gives Minorities a Chance to Make Their MARC

    1. Jeffrey Mervis

    Should a program aimed at boosting the number of minority scientists target minority students? NIH thinks so, and the Supreme Court seems to agree

    The U.S. Supreme Court was about to rule on whether colleges could use race-conscious admissions policies when the directors of the undergraduate Minority Access to Research Careers (MARC) program gathered last month in California for their annual meeting. Many wondered if they were living on borrowed time.

    They knew that the MARC program, sponsored since 1982 by the National Institutes of Health (NIH), is possibly the last remaining federal effort to use race as a factor in determining eligibility. They also knew that its approach to increasing the number of minority scientists seemed out of step with the Bush Administration's public opposition to affirmative action and with the arguments it had filed with the court. As a result, many of the MARC directors worried that NIH might decide to reconsider its support for the $25-million-a-year program if the high court struck down the University of Michigan's policies.

    But a week later, after returning to their universities, the 54 MARC directors breathed a sigh of relief. The high court had just ruled that race is a legitimate factor in admissions policies aimed at fostering campus diversity (Science, 27 June, p. 2012), a stance that many took as a vote of confidence in their efforts, too. “We could have been set back considerably if the decision had gone the other way,” says biochemist Elma Gonzales, who coordinates the MARC program at the University of California, Los Angeles. “I'm not sure if we'll expand. But the need is still there. And the court's ruling keeps the door open.”

    Indeed, MARC is all about opening doors. Each year, it gives 700 talented undergraduates an intensive introduction to the life of a scientist, subsidizing their education, putting them to work in the lab, and offering them one-on-one career counseling. For Nanibaa Garrison, born and raised on a Navajo reservation, MARC provided the opening for a biology degree this spring from the University of Arizona in Tucson and a chance to head off this fall to a doctoral program in human genetics at Stanford University. It's also nurtured her dream of returning to the reservation someday to do research that could improve the health of all Native Americans.

    A life in the lab.

    Nanibaa Garrison says that the MARC program has sold her on a research career.


    Like other federal and private efforts designed to benefit underrepresented minorities, MARC has come under scrutiny in recent years. Officially, the program is not race-exclusive, with NIH allowing institutions to set their own criteria for eligibility. “MARC is not limited to minorities,” says Clifton Poodry, who oversees MARC and other minority training programs at the National Institute of General Medical Sciences.

    At the same time, several MARC directors say that NIH expects them to recruit students from the groups designated as underrepresented in science: African Americans, Hispanics and Latinos, Native Americans, and Pacific Islanders. “After all, the bottom line for NIH is how many underrepresented minorities pass through the program and go into Ph.D. programs,” says Victor Rocha, MARC director at California State University, San Marcos. The program's Web site ( tiptoes around the issue. It notes that the awards provide support “for students who are members of minority groups underrepresented in the biomedical sciences,” adding cryptically that these groups “include, but are not limited to,” the four federally designated groups.

    In practice, all but a few of the 54 MARC institutions restrict participation to the four groups. As a result, NIH estimates, more than 90% of the roughly 700 MARC scholars in any given year are members of those groups. The most notable exception is the University of Maryland, Baltimore County (UMBC), which coincidentally has the largest MARC program in the country, where majority students from disadvantaged backgrounds typically fill about half a dozen of its 40 slots.

    UMBC officials acknowledge that it wasn't their idea to open the program. But a 1996 federal court decision striking down racial criteria for another University of Maryland scholarship program for minorities forced their hand, explains Lasse Lindahl, chair of the biological sciences department and MARC director. “When we got the MARC grant in 1997, the state attorney general told us that we needed to follow the same rules,” says Lindahl.

    Garrison's dream of setting up a modern biology lab on the Navajo reservation grows out of her positive research experiences in the MARC program, she says. And her decision to pursue a doctoral degree “probably puts me ahead [academically] of everybody else in my high school graduating class.” Her career goal also makes her a poster child for why MARC supporters believe a race-exclusive training program is important. “Researchers are attracted to topics that affect their lives. And minority scientists are more likely to work on issues that affect minority health,” says biochemist Marc Tischler, who runs the University of Arizona's MARC program.

    View this table:

    MARC directors have wrestled with the same issue, and most have come down in favor of race as the determining factor. “There are lots of other opportunities for majority students on this campus who want to do research,” says Tischler. “This program is specifically designed for minorities.” Mathematician Joaquin Bustos, who directs the MARC program at Arizona State University in Tempe, worries that some underrepresented minority students would be pushed aside if the program were opened up to everybody. “Many minority students aren't as prepared to apply for things like MARC,” says Bustos. “Majority students are more likely to say, ‘Hey, this is a great opportunity.’”

    Still, a few MARC directors say that the program might benefit from broader eligibility standards. “It's important for students to hear a diversity of views,” says health psychologist Cathie Atkins, an associate dean at San Diego State University, who notes that the university's MARC program includes workshops that are open to non-MARC students. “I'd like to see NIH go to a three-pronged standard [that includes socioeconomic factors],” she says, “so long as MARC students still demonstrate a desire to look at the racial disparities in our health care system.”

    Poodry says that NIH officials have resisted suggestions over the years to modify the MARC criteria to focus on low socioeconomic status rather than race. “The problem was, compared to whom?” he asks. “We could never figure out how to craft a definition of ‘disadvantaged’ that would work.”

    NIH has never figured out how to tell if MARC is delivering on its promise, either. Institutions submit reports on how many MARC scholars graduate, but Poodry admits that the material has never been analyzed. The long-term goal of turning MARC students into applicants for NIH's bread-and-butter R01 awards is even more elusive. “We have anecdotes but no hard data,” says Poodry. A 1995 study “fell short of our expectations,” he says, “and Congress has never asked us for that type of scorecard.” Two years ago, NIH gave the National Academy of Sciences $1.5 million to review 20 programs aimed at boosting the number of underrepresented minorities, including MARC and two other undergraduate programs. But it will be at least another year before the academy issues its report.

    Whatever the academy concludes about the effectiveness of the MARC program, most directors already believe fiercely in the importance of trying to attract more minorities into science. “I don't apologize for running a program that serves minorities,” says Robert Koch, a cell biologist at California State University, Fullerton, and longtime MARC director. “I have a personal commitment to this program, and I feel that this is the only way we are going to make progress.” Adds Rocha, “Even if NIH backed away [from MARC], we'd still try to attract more minorities into science.”


    Scientists Dream of 1001 Complex Mice

    1. Gretchen Vogel

    Geneticists want to create 1000 new lines of inbred mice to help them sort out the genes behind complex diseases such as diabetes and cancer

    OXFORD, U.K.—As any lover of fine Scotch will tell you, a little complexity is a good thing. The trick for the distiller is to produce a complex, but consistent, product. Geneticists trying to sort out the genes that trigger some of medicine's most common ailments face a similar challenge. They need to capture the natural variation that exists in human populations but in a way that is reproducible in the lab.

    Despite the sequencing of human, mouse, and other genomes, researchers are a long way from understanding how genes contribute to most common ailments, from depression to diabetes to skin cancer. Sorting out the tangle of genetic and environmental factors that lead to such maladies is one of the biggest challenges of the postgenome era.

    Already, several multimillion-dollar projects—including the $110 million human HapMap (Science, 24 May 2002, p. 1391)—are under way to help scientists tease apart the intricate connections. Now a group of mouse geneticists is poised to add a new megabiology effort. They are hoping to win support—and at least $50 million over 10 years—to create as many as 1000 new lines of inbred mice.

    The geneticists propose to cross eight commonly used mouse strains for four generations, producing litters with all 1680 possible permutations of great-grandparents. They will then turn to inbreeding, performing brother-sister matings until the offspring of each strain are essentially clones. Not all strains will survive the inbreeding, but after 20 generations—about 7 years—the scientists hope to end up with a palette of roughly 1000 different strains. Together, these so-called recombinant inbred (RI) lines would reflect much of the genetic variation present in natural populations. Yet each individual line would have the advantage of being easily reproduced, cookie-cutter fashion, in exact copies. “Mouse genetics has been focused on making things simpler. We want to make things more complicated,” says geneticist Robert Williams of the University of Tennessee Health Science Center in Memphis, one of the leaders of the RI proposal.

    Supporters of the RI lines say they would significantly speed the path to specific genes as well as uncover new interactions among genes. Skeptics agree that a new approach is needed but caution that the logistics and expense of producing such a resource may outweigh its benefits.

    Identical quintuplets.

    Inbred mice such as these from the A/J line can help researchers track down the genes behind complex traits.


    At a meeting here in early July,* several dozen statisticians, geneticists, and other biologists gathered to discuss the RI project and other methods for tracking down quantitative trait loci (QTLs). These are regions of the genome that seem to play a role in complex traits such as weight gain or susceptibility to cancer. Although scientists using a variety of methods have been fingering increasing numbers of QTLs in both mice and humans, getting from general locale to specific gene is much more difficult. The successes—for example, the connection between depression and the serotonin transporter gene (Science, 18 July, p. 386)—have been hard won.

    One of the factors confounding the search in humans is that people who carry a susceptibility gene may also carry other protective genes, masking the effect of the disease-prone allele in epidemiological studies. “You can't study complex genetics the way you'd like to in humans,” says geneticist Gary Churchill of the Jackson Laboratory in Bar Harbor, Maine. “You can't develop inbred lines, and you can't arrange matings.”

    To date, most studies that use mice and other animals to discover the role of genes have relied on two standard genetic techniques, mutagenesis and traditional crosses. In mutagenesis, scientists expose mice to chemicals or radiation that disrupt DNA and then look for interesting traits in the offspring. Such studies are very effective for elucidating the role of single genes; indeed, international researchers are working to create a stock of mice with a mutation in every single gene (Science, 2 June 2000, p. 1572). But these studies are less useful for understanding how multiple genes work together, Churchill says. “Natural variation is fundamentally different from mutagenesis. Diabetes doesn't arise because of a point mutation. It's a genetic background that interacts with the environment.”

    Controlled complexity.

    To help track down genes involved in complex traits, scientists want to create 1000 new lines of “recombinant inbred” mice. Each line would be a combination of eight existing strains; the scientists would outcross the mice for four generations and then breed brother-sister pairs for 20 generations to create inbred lines.


    To get at more-complex traits, scientists can cross existing strains of mice, screen the offspring for traits of interest, and then look for genetic regions that the affected animals share. For example, a scientist looking for genes related to obesity can cross a mouse from a heavy strain with one from a more svelte line and then, after several generations, search for genes that the heavy cousins have in common. But the problem with such crosses is that they are unique events. Like any wild mating, the assortment of genes in the offspring is random, and once the mice die, their particular genetic mixture can't be recreated. “You measure the mice, and poof, they're gone. If there's an interesting one, you can't go back and put it on a high-fat diet,” Churchill says.

    RI lines offer a solution. Scientists will be able to quickly trace the origin of any DNA segment in an RI line back to one of the original eight founder strains, speeding the search for genes shared by lines that have a particular trait. And by simply breeding any male-female pair from the desired line, researchers can create more mice with exactly the same mix of genes. Because of the way the RI lines are derived from the original eight-way cross, each line would be at least a second cousin to all the other RI lines. That means that researchers widely separated in space or time could experiment on matching sets of cousins, exposing them to different diets, cancer-causing chemicals, or mental stresses, for example. The results of such screens would be cumulative for each line.

    Indeed, the project's proponents want to build a publicly accessible database in which to collect experimental results from around the world so that a neuroscientist in the Netherlands, say, could build on the work of an endocrinologist in British Columbia. As data accumulate on the behavior of different strains in response to different conditions, scientists will be able to run ever more sophisticated and powerful computer searches for the genes that play a role in disease-related traits. “If there's some connection between bone density and blood pressure, then the evidence will be there,” Churchill says.

    Williams and Churchill are part of an unofficial group, the Complex Traits Consortium, that is interested in tracking down the variations behind multigenic traits. Along with several colleagues, they hatched the idea of creating hundreds of RI lines at a meeting a year ago; they have since refined it based on input from the consortium's e-mail list server. Early results with more than 100 existing RI lines—developed by Williams and at several other laboratories—are promising. At the Oxford meeting, Leonid Bystrykh and Gerald de Haan of the University of Groningen, Netherlands, described how the strains had enabled them to identify several genes that affect the proliferation of blood stem cells in mice and are apparently linked to several types of leukemia in humans.

    But the project's price tag—estimated at $50 million—gives some people pause. “Everyone would love it if such a resource existed,” says Mark Daly, a geneticist at the Massachusetts Institute of Technology. “But there are different ways of going about [finding QTLs]. … Any project of this magnitude requires a great deal of justification and careful thought.”

    Churchill, Williams, and the other proponents agree that the idea needs some fine-tuning. Before they put together a formal proposal for funders, they say, they need to decide which lines to use as parental strains, exactly how many lines they ultimately want to make, and how and where the resulting lines will be stored and distributed. But Churchill is impatient. In the meantime, he says, “we're going to start making some RIs in our lab. Once the population reaches a certain size, people will begin to understand what the technology can do.”

    At this stage, argues Joseph Nadeau of Case Western Reserve University School of Medicine in Cleveland, mouse geneticists should be exploring several options for finding QTLs. He and his colleagues, for instance, are carefully breeding mice so that a single chromosome from one inbred strain is substituted in the genome of another strain. “The more resources we have, the better,” agrees Churchill. The trick, they both concede, will be finding enough money to support such exquisite complexity.


    Male Bugs Sponge Off of Their Mates

    1. Elizabeth Pennisi

    CHICO, CALIFORNIA—The evolving understanding of how bizarre bugs, invasive weeds, and many other flora and fauna arose dominated conversations among the 1400 biologists gathered here 20 to 24 June.

    Few modern women consider it their duty to take charge of their husband's nutrition. But females of a newly discovered insect called the zeus bug take full responsibility for their mate's meals, particularly during the mating season. It's a case of role reversal in the entomological arena and “the first example I know of” in which females are the providers, says evolutionary biologist Edward Morrow of the University of California, Santa Barbara.

    In dozens of other insect species, males court female partners with “nuptial” gifts. Some offerings are the equivalent of Valentine's Day candy: tasty, nutritious morsels that energize the female and make her look favorably upon her suitor. In some instances, the gift keeps her from trying to chow down on her mate. Other gifts contain chemicals that are toxic to potential enemies and help protect the eggs. In all cases, “the males are passing on to the female a paternal investment,” says evolutionary biologist Göran Arnqvist of the University of Uppsala, Sweden. He reported these findings at the meeting and in the 24 July issue of Nature.


    Female zeus bugs provision their piggybacking mates.


    Arnqvist and colleagues Therésa Jones and Mark Elgar of the University of Melbourne, Australia, have now demonstrated that females can also be donors. They studied one of four species of the Phoreticovelia genus discovered by other researchers 4 years ago in New Guinea and tropical Australia. Close cousins to water striders, zeus bugs run along the surface of the water, scavenging for insects and other prey. But during their field studies, the researchers discovered that unlike water striders, males often rode piggyback on female nymphs and that this practice continued into adulthood. Even though mating was brief, males usually stayed glued to females for the entire 2- to 3-week mating season.

    Arnqvist was intrigued. He observed that both mature and immature females have an unusual gland on their backs—right where the male's head is positioned as it hangs on. To find out whether the female was rewarding her companion with food, the researchers fed females radiolabeled fruit flies. They followed the meal's radioactive signal from female to male, showing that the males were feeding off of the females, Arnqvist reported at the meeting. The experiment “shows a direct transfer of nutrients to the male,” Morrow says.

    Moreover, the bonus nutrition did a world of good for the male: “If the male sits on a female, then he doubles his life span,” Arnqvist reported. He suspects that the female's bonus is that the males don't try to eat her. Moreover, as burdensome as the male might seem to the female, having a constant companion didn't seem to affect her that much: By removing some males from the females after mating, the researchers demonstrated that females produced about the same number of eggs and lived about as long regardless of whether they carried a passenger.

    Even so, the females' generosity was “kind of puzzling,” Arnqvist noted. It usually doesn't make sense for the females to be the gift-givers. By producing eggs, which require more energy than sperm, they're already doing more than their share of the heavy lifting in the relationship. And what is even more perplexing is that the male isn't always a gentleman in response to her generosity. Some male-sporting females had scars along their backs, occasionally more than 50 punch marks. “We have to ask whether the males are stinging the females or even cannibalizing them,” Arnqvist suggested.


    Garden Flower Evolves Into Weed

    1. Elizabeth Pennisi

    CHICO, CALIFORNIA—The evolving understanding of how bizarre bugs, invasive weeds, and many other flora and fauna arose dominated conversations among the 1400 biologists gathered here 20 to 24 June.

    Some nonnative plants docilely flower only in carefully tended gardens. Others are more overbearing, sneaking into new territories and crowding out native species. Now evolutionary biologists are uncovering what separates the tame from the wild: Within a given alien species, some individuals may be more genetically suited to wreak havoc than others. Given the right conditions, this small subpopulation can overrun a new ecosystem. The study, presented at the meeting, implies that any exotic plant has the potential to become invasive.

    Alien plants, animals, and pathogens cause more than $130 billion per year in damage in the United States alone. Some were stowaways on ships or airplanes; others are descended from escaped pets or were brought in for landscaping before researchers recognized their invasive powers.

    In Canada and the northeastern United States, a perennial called Silene latifolia, or white campion, is spreading and interfering with alfalfa and other grain crops. But in its native Europe, it is much better behaved, confining itself primarily to gardens and roadsides. In North America, says Amy Blair of Georgia Southern University (GSU) in Statesboro, “we have an invasive phenotype that is born to run.”

    Blair and Lorne Wolfe, also of GSU, are among a new breed of biologists scrutinizing invasive plants. Researchers in the past have concentrated on the ecological consequences of invasive species. They have identified factors conducive to the establishment of these intruders and qualities of these species that enable them to cope with a variety of conditions. But this new group of researchers is investigating the role of evolution and genetic change in creating biological monsters.

    Weed genes.

    The invasiveness of white campion in North America has genetic underpinnings.


    To test whether evolution was at work in white campion's spread, Blair and Wolfe collected seeds from 40 sites across Europe and North America and grew them in a greenhouse for a few weeks. Then they transferred 400 plants to a field station in Virginia and monitored their growth.

    The seeds from the United States and Canada took off. A greater proportion sprouted, and on average plants emerged from the soil 2.5 days earlier than did those from European seeds. North American progeny were bigger, had more leaves, and flowered earlier. More of them survived the winter. “There has definitely been evolution in plant characters that are associated with weediness or invasiveness,” Wolfe says.

    The European plants did outdo the North American ones in one respect, however: They produced more protective hairs called trichomes. That difference makes sense, says Wolfe, because “there are specialized enemies in Europe that are not here.” This finding fits with the notion that in new environments, alien species have fewer predators and thus can devote more energy to growing and spreading.

    Blair says she has yet to tease out exactly what's gone on in white campion. It could be that only the more aggressive plants were able to establish themselves on North American soil, and they have passed those traits down to the plants the team studied. Interbreeding between native plants and the intruder might have produced the right mix of traits for rapid colonization. Or natural selection may have come into play: Only once certain genetic mutations accumulated did the plant become so aggressive. Native species are less likely to evolve such characteristics because pathogens and other enemies typically keep them in check.

    If the latter explanation in particular is correct, then “the information from this study has practical significance,” says Heather Davis, an evolutionary biologist at the University of California, Davis. “It raises an alarm bell about ‘naturalized’ species that aren't considered a threat but may in fact be sleeping giants, quietly evolving until they overcome some barrier to explosive population growth.”


    Genetic Deficiency Blamed on Bacteria

    1. Elizabeth Pennisi

    CHICO, CALIFORNIA—The evolving understanding of how bizarre bugs, invasive weeds, and many other flora and fauna arose dominated conversations among the 1400 biologists gathered here 20 to 24 June.

    Among certain organisms, females seem to have won the battle of the sexes—or at least a tussle over DNA allocation. In about 20 groups of insects, mites, nematodes, and invertebrates called rotifers, females have chromosomes that come in pairs, whereas their male companions have to make do with just a single copy of each one.

    For decades, biologists have argued about how this unequal way of life, called haplodiploidy, could have evolved. Now Benjamin Normark, an evolutionary biologist at the University of Massachusetts, Amherst, argues that in the 10 groups of haplodiploid insects, at least, microbes ganged up on the males. His evidence comes in part from studying the literature about the phylogenetic relationships, evolutionary ancestries, and early life histories of groups of insects with such lopsided genetics. The work “allows him to infer the factors involved in the several origins of haplodiploidy,” says Peter Cranston, an evolutionary biologist at the University of California, Davis.

    Haplodiploid insects share certain characteristics. Most feed on dead or living woody plant stems, Normark says. Often, a female lives crammed together with her young, for example, in a cubbyhole under bark. However, what's key, he claims, is that their diet requires them to enlist the digestive capacities of a variety of microbes.

    Normark wasn't the first to theorize about these commonalities. In 1967, evolutionary biologist William D. Hamilton tallied more than a half-dozen features shared by haplodiploid insect species. A dozen years later, he suggested that inbreeding resulting from living in close quarters under bark helped select for this genetic system. And in 1993, he proposed that microbes called endosymbionts might play a role, but he never explained how.

    Instead, Hamilton concentrated on explaining why haplodiploidy exists. He decided that the system evolved as a way for an inbreeding female to control the sex ratio of her young. In such cases, the more daughters she could produce, the bigger her family—a single male would do to fertilize all her daughters' eggs. Other evolutionary biologists thought haplodiploidy evolved as the female's one-upmanship for passing on her genes in situations with more outbreeding. The diploid female of the species gets more of her genes into the offspring than does the haploid male. The matter remained unresolved.

    That's where Normark comes in. “I am suggesting that all of Hamilton's observations fit together,” Normark says. His scenario can even encompass the other researchers' ideas, he adds. Normark lays most of the blame for haplodiploidy on endosymbionts, microbes that live inside cells of their host. He speculates that these microbes first served the insects by digesting wood products and later evolved a less altruistic ability: to manipulate reproduction in the host to their own advantage.

    A crucial piece of evidence for this theory is still missing: No one has surveyed the haplodiploid community to find out which endosymbionts are making their mark in these species. But based on studies of an endosymbiont called Wolbachia, researchers know that these organisms can take a toll on the production of males. Why go after males? Endosymbionts sneak into the next generation stowed away in eggs—sperm don't carry the microbes—so the more female offspring a host produces, the more avenues of infection are opened for these microbes. The female benefits, too, as she wants to skew the sex ratio in her favor, but the microbes drove the evolution of this odd genetic system.

    Normark proposes that the microbes these insects enlisted to help digest their food somehow evolved ways of getting rid of half the male's chromosomes—a move that initially killed most of the males. But over evolutionary time, the males that could survive with depleted chromosomes managed to withstand the microbes' assaults, and haplodiploidy became a way of life. Once established, it persisted even as new species—including bees and their relatives, in which males are likewise endowed with haploid genomes—forsook their wood cubbyholes and developed new lifestyles and food sources.

    Normark plans to examine haplodiploid scale insects to see if their life history and endosymbiont load are in line with his theory. And his colleagues think his ideas may help shape other studies of haplodiploidy. “This is a long-standing problem, and he makes a compelling case for having solved it,” says Gerald Wilkinson, an evolutionary biologist at the University of Maryland, College Park.