News this Week

Science  02 Jul 1999:
Vol. 285, Issue 5424, pp. 18

    Top Official Resigns as Congress Pushes for Management Changes

    1. David Malakoff

    The controversy over allegations of Chinese spying at U.S. nuclear weapons laboratories claimed its first high-level victim last week with the resignation of Victor Reis, head of the Department of Energy's (DOE's) nuclear weapons program. Reis had clashed with Energy Secretary Bill Richardson over a Senate proposal to revamp oversight of the agency's bomb-making enterprise, which includes laboratories that conduct both military and civilian research. But Reis's departure does little to clarify the status or ultimate fate of the labs and the thousands of scientists who work there.

    Reis reportedly supported a sweeping plan, which the Senate could approve as early as this week as part of an intelligence bill, to create a new, independent weapons agency within DOE. Richardson strongly opposes the scheme, calling it a bureaucratic nightmare. The reorganization also faces opposition in the House, where some lawmakers worry that it could weaken the quality of DOE's civilian science programs. In the meantime, Richardson and others have successfully rebuffed an effort to severely restrict DOE's foreign visitors program, which brings thousands of overseas scientists to the labs.

    The spying controversy erupted in March after The New York Times publicized allegations against Wen Ho Lee, a physicist at DOE's Los Alamos National Laboratory in New Mexico (Science, 7 May, p. 882). The charges figured prominently in a massive report released last month by a special House committee investigating Chinese espionage in the United States. Lee, who was fired in April for security breaches, has denied giving nuclear warhead data to China.

    DOE critics won a round earlier this month when a special White House panel led by former Republican Senator Warren Rudman of New Hampshire issued a report calling for a massive organizational shake-up. “For the past two decades, [DOE] has embodied science at its best and security of secrets at its worst,” the panel concluded, calling the agency a “dysfunctional bureaucracy that has proven it is incapable of reforming itself.” Among its recommendations was shifting DOE's weapons work into a new, semiautonomous agency led by an experienced national security expert.

    The report galvanized Republican senators Pete Domenici (NM), Jon Kyl (AZ), and Frank Murkowski (AK), who quickly introduced legislation to implement its suggestions. Their proposal would create an Agency for Nuclear Stewardship to oversee the agency's sprawling weapons complex, including the three nuclear labs and about a dozen other bomb-making sites. It would be led by a high-ranking DOE official reporting only to the secretary, with its own budget and hiring authority, and freedom from DOE oversight on security and environmental issues.

    Such independence makes the idea unacceptable to Richardson and other critics, who would prefer a smaller management tweak: the addition of a high-ranking official who would oversee the weapons program. At a 23 June hearing before the Senate Armed Services committee, Richardson complained that the measure “undermined chains of command” and that placing the labs' nonweapons research—such as solar energy and climate studies—under the control of a security agency “would be a disaster.” In other comments, he suggested that it would erect a “Berlin wall” around the labs' nonweapons work.

    In another hearing last week, Representatives Tom Bliley (R-VA) and John Dingell (D-MI), leaders of the powerful House Commerce committee, voiced doubts about the Senate plan. “We need to ensure that we don't trade old problems just to find ourselves with new ones,” said Bliley. Other skeptics include Representatives James Sensenbrenner (R-WI) and George Brown (D-CA), leaders of the Science committee, which was set to review the Rudman report on Tuesday. “There are going to be a lot of questions about how civilian science would fare under the security thumbscrews,” predicted a House staffer.

    Reis, however, told Richardson that he favored the Senate proposal, according to one Senate aide. A Bush administration appointee, Reis is credited with fostering the agency's $4 billion-a-year Stockpile Stewardship Program. Its goal is to use powerful lasers and supercomputers to simulate nuclear explosions, replacing the tests that were outlawed under the Comprehensive Test Ban Treaty, which the U.S. signed in 1996 but has not ratified. But some legislators have faulted Reis for failing to inform Congress or the White House of the spying allegations. His support for restructuring the department was the last straw for Reis, according to the Senate aide. Science could not reach Reis for comment, but Domenici, Kyl, and Murkowski have accused Richardson of “forcing” Reis out.

    The legislative jousting did bring a positive development for science: the defeat of a proposed curtailment of foreign visitors. On 9 June the House overwhelmingly rejected a proposed 2-year moratorium on the visitors program, opting instead for a 90-day time-out if DOE cannot show it has strengthened visitor background checks. Earlier, the Senate approved a similar plan, although its ultimate legislative fate remains uncertain. In the meantime, DOE officials are watching closely as Congress sets a course that could decide DOE's shape for years to come.


    Has a Cosmic Standard Candle Flickered?

    1. James Glanz

    A team of astronomers says it has found slight, previously unnoticed variations among the exploding stars called type Ia supernovae. These explosions, thought to flare up to roughly the same brightness each time, have served the crucial role of cosmic “standard candles” whose apparent brightness, as seen from Earth, can serve as a measure of their distance. Unexplained variations in the explosions could, in theory, call into question the cosmic measurements based on them—among them the dramatic finding that the expansion rate of the universe is speeding up over time (see Science, 18 December 1998, p. 2156).

    The discrepancies emerged when Adam Riess, Alexei Filippenko, and Weidong Li of the University of California, Berkeley, and Brian Schmidt of Mount Stromlo and Siding Spring Observatory in Australia looked closely at the early phase of 10 nearby type Ia explosions. The team found that the time it took the explosions to reach their peak brightness was more than 2 days longer than the average for supernovae billions of light-years away as measured by a second group. “They're pretty strongly discrepant with one another,” says Riess of the two data sets. “If it's true, it's extremely interesting,” says Eddie Baron, an astrophysicist at the University of Oklahoma, Norman, who saw Riess's presentation at a workshop in Aspen, Colorado, on 17 June.

    Astronomers on the Supernova Cosmology Project, the team that studied the distant supernovae, emphasize that they have not completed their own analysis, let alone published it. “I don't even know if [the difference] is going to hold water,” says Berkeley astronomer Peter Nugent, a member of that team. And no one knows what, if anything, the difference in the supernovae's rise times might mean about their value as standard candles. “It's just an illustration that there is a bit more going on than in the simplest … models,” says Craig Wheeler, a theorist at the University of Texas, Austin.

    Type Ia supernovae are prized as distance indicators not only because they seem to explode in nearly the same way each time, but also because astronomers can account for leftover brightness differences. Studies of supernovae at the same distance have shown that the rise and fall of brightness, which unfolds over several months, takes longer for slightly brighter explosions. But because the explosions are generally not spotted until they are well under way, astronomers had never examined in detail the interval between a supernova's appearance and its peak.

    Now, Riess and his colleagues have filled that gap by drawing on a robotic telescope that Filippenko and others operate, as well as on observations by a team of supernova watchers at the Beijing Astronomical Observatory and by amateurs. Indeed, Riess says that Chuck Faranda, an amateur astronomer from Florida, spotted the freshest explosion using an electronic camera hooked to a small telescope in his back yard. The Supernova Cosmology Project, whose analysis was led by Berkeley's Gerson Goldhaber, relied on surveys of many galaxies to spot large numbers of distant supernovae, catching some of them early in their history.

    The difference between the two sets of measurements, if it is meaningful, might ultimately provide a new “calibration” to correct for residual brightness differences. In the meantime, however, the possibility that distant supernovae are intrinsically different from nearby ones could raise questions about the most celebrated use of type Ia supernovae: studies of how the expansion rate of the universe has changed over time. Last year, the Supernova Cosmology Project and another group, called the High-z Supernova Search Team (of which Riess, Filippenko, and Schmidt are members), found that distant type Ia's were unexpectedly dim. Providing the explosions are truly standard, the dimness implies unexpectedly great distances caused by an acceleration of the expansion over billions of years. Despite the astronomers' best efforts, they have found nothing to challenge the conclusion until now.

    “We're in a very early state,” says Saul Perlmutter of Berkeley, leader of the Supernova Cosmology Project. “There is so much uncertainty regarding this result that I don't either believe or disbelieve it,” adds Philip Pinto, an astrophysicist at the University of Arizona, Tucson, who also heard Riess's talk. But Pinto is pleased that astronomers are examining type Ia's from every possible angle in order to test their performance as standard candles. Says Pinto, “That's science working as it should.”


    Rubber Mold Carves a Path to Micromachines

    1. Alexander Hellemans*
    1. Alexander Hellemans is a writer in Naples, Italy.

    Photolithography, the chemical printing process used to make the circuits of computer chips, has allowed manufacturers to shrink their devices to almost unimaginably small sizes. But while the process can easily handle the two-dimensional structures in an electronic circuit, it has much more difficulty with 3D structures like cavities and microchannels, the sort of things required by the new generation of micromechanical devices, such as transducers that capture sound waves, as well as miniature chemical plants and “labs-on-a-chip.” Now, a team of researchers at Harvard University in Cambridge, Massachusetts, reports on page 83 that it has developed a technique for fashioning microstructures in 3D with the help of liquids passing through a network of channels, exploiting the flow patterns to deposit or etch away structures in layers of metals or other substrates, including crystals, ceramics, or organic polymers.

    Photolithography makes poor 3D structures because it builds up many thin layers one at a time through a complicated process of shining light through a mask onto a photosensitive chemical on the surface, dissolving the light-exposed regions so that a pattern of bare surface is revealed, and then etching away or chemically treating the bare surface. This process can be repeated many times to build up devices, but creating fully 3D structures demands extremely precise alignment of the masks. The Harvard team, led by materials scientist George Whitesides, does away with the need for masks and alignment by defining the shape of the structures with a pattern of capillary channels pressed onto the surface.

    The key to the technique is polydimethylsiloxane—otherwise known as silicone rubber. “The big advantage of that polymer is that it will come into tight contact with most surfaces,” says team member Rustem Ismagilov. He and his colleagues create patterns of grooves in the surface of the silicone rubber by polymerizing the rubber sheet on a master with ridges on its surface, similar to the way vinyl records are made. Then they press the rubber sheet onto the flat substrate to create closed capillary channels. By passing different chemicals through these capillaries, the researchers can etch away the surface of the substrate or deposit material onto it, following the pattern marked out in the silicone rubber.

    The researchers found that they could also deposit material at specific points within a capillary, creating features as small as 3 micrometers, which Ismagilov says does not compare badly with the 0.1 micrometer now possible with photolithography. They relied on laminar flow, a turbulence-free state that develops in fluids under certain conditions. “At the sizes of capillaries we have, it is almost impossible to create a flow that is not laminar,” says Ismagilov. As a proof of principle, the researchers exploited laminar flow to deposit silver not across the whole width of the capillary, but just in a narrow strip down the middle.

    They introduced the two components of a commercial silver plating solution as two parallel flows in a zigzag-shaped capillary. Because there was no turbulence, the two solutions flowed side-by-side without mixing. They reacted only at their interface, depositing a thin silver thread on the bottom of the capillary. The team went on to use the technique to create a three-electrode microelectrochemical detector inside a capillary: First, they deposited a gold strip on a surface, then etched away a stripe down the middle of it to form two electrodes, and, finally, deposited a silver reference electrode in between the two gold electrodes. Whitesides now has his sights set on making several other types of devices, such as very small detectors and light sources. “I'm hopeful that we can get these systems to lase,” he says.

    Marc Madou, a microfabrication researcher at the Ohio State University in Columbus, calls the technique “elegant.” Both he and Whitesides agree that the technique does not have a great future in high-volume manufacturing because it requires intensive monitoring, for example, of the flows in the capillaries. But, Madou says, it is a “good laboratory tool” for making small experimental devices used in a wide range of research fields, including chemical and biochemical analysis and electrochemistry.


    A Good SNP May Be Hard to Find

    1. Michael Hagmann

    Over the past 2 years, academic and corporate labs alike have been swept up in a human DNA gold rush. They have eagerly mined the human genome for minute differences between individuals, hoping to use the information to analyze common diseases and create powerful, custom-made drugs. The target: single-base variations in DNA—or single-nucleotide polymorphisms (SNPs)—that occur about once in every 1000 bases of the 3 billion bases in the human genome. Many researchers hope that random collections of these mutations will yield a shortcut to identifying the genes underlying such major diseases as asthma or cancer. But now, findings by a couple of major labs in this field suggest that the payoff of this strategy will not come any time soon, because the most common type of SNPs may not be the most informative.

    This cool appraisal comes from two leading teams in SNP research, one headed by human geneticist Aravinda Chakravarti of Case Western Reserve University in Cleveland, Ohio, and the other by Eric Lander, director of the genome center at the Whitehead Institute for Biomedical Research in Cambridge, Massachusetts. Both groups published reports in the July issue of Nature Genetics based on SNP collections they gathered from about 200 different human genes. Their analyses suggest that a popular approach to SNP hunting—comparing entire genomes of just a few individuals to find random variations—may miss most of the SNPs that alter the structure of the proteins they encode. Yet these are the SNPs that may directly influence disease risk.

    Most SNPs, according to Chakravarti, are not likely to have a direct impact on their protein products. This is because they fall in the estimated 95% “noncoding” area of the genome, or because they behave in a “synonymous” or silent way, coding for the same protein an alternate SNP codes for. Nonsynonymous coding SNPs (cSNPS), in contrast, are very rare in the human gene pool. “There seems to be a strong selection against any change in protein structure. [Most of these changes] have been weeded out in the course of evolution,” says Chakravarti.

    In addition, Lander's study reports that a significant percentage of the relevant cSNP variants are found mainly in certain subpopulations, such as Asians or African-Americans. “What that tells us,” says Leonid Kruglyak, a geneticist at the Fred Hutchinson Cancer Research Center in Seattle, “is that the [nonsynonymous cSNPs] are harder to find in the first place.” Chakravarti agrees, adding that “to discover them you'll have to take as large and diverse a sample population as possible.”

    The scarcity of protein-altering SNPs will also make it difficult to tie them to a specific disease. Lander's study concludes that linking a disease to a very rare gene variant would require thousands of patients, way too many even for state-of-the-art tools for whole genome analysis. At present, says Chakravarti, “The right way to go is to take a set of candidate genes and assess them directly in as many patients as possible” for an association between SNPs and the disease.

    This cautionary advice comes as the SNPs stampede is well under way. In January 1998, for instance, Francis Collins, director of the National Human Genome Research Institute (NHGRI) at the National Institutes of Health in Bethesda, Maryland, launched a $30 million project to create a collection of some 450 human DNA samples that aims to expand the number of SNPs from a few thousand known today to about 100,000 in the next 3 years. And in April, 10 large drug companies, the Wellcome Trust philanthropy of Britain, and a handful of academic laboratories teamed up to form a nonprofit alliance called the SNP Consortium, or TSC, that will create a SNP archive encompassing some 300,000 SNPs within the next 2 years. Like J. Craig Venter's sequencing factory Celera Genomics in Rockville, Maryland, TSC will collect random data across the entire genome.

    In combination, the gene-focused and random strategies for collecting SNPs should enable scientists to explore the human genome extensively. Celera geneticist Mark Adams says: “I really see [the two approaches] as complementary. The whole-genome strategy will give us a large number of SNPs, both within and outside of genes, and that's a very useful starting point.” With this large SNP bonanza just ahead, says Lisa Brooks, program director of the SNP project at NHGRI, research sponsors need to develop new automated technologies that can rapidly “score the genotypes of many individuals.” NHGRI is putting money into this area. But Brooks notes that “even a standard screening of, say, 50 candidate genes is still a way off.”


    OECD to Set Up Global Facility on Biodiversity

    1. Judy Redfearn*
    1. Judy Redfearn writes from Bristol, U.K.

    Researchers studying the diversity of life have dreamed of pooling all they know in a single electronic compendium. Last week in Paris that dream moved a step closer to reality when the science ministers of 29 industrial countries agreed to create a Global Biodiversity Information Facility (GBIF).

    The virtual facility hopes to convert a growing tower of biodiversity Babel, replete with incompatible databases, confusing terminology, and uncatalogued material, into a transparent source of information that is accessible to anyone, anywhere. But before it tackles that challenge, GBIF will have to be transformed from an attractive notion into a real facility with a staff and a budget.

    GBIF is the fruit of a 2-year effort by a working group of the Megascience Forum, a body created by the industrialized-country members of the Organization for Economic Cooperation and Development (OECD) to explore possible collaborations in building large scientific facilities. Last week the science ministers agreed it should be renamed the Global Science Forum to better address scientific issues of global significance that do not necessarily involve major construction—such as biodiversity. Last year the parties to the 1992 United Nations Convention on Biological Diversity (CBD) urged OECD to come up with a program like GBIF to give individual countries access to the scientific information they need to carry out the terms of the convention. “GBIF is a very important international undertaking to ensure we can all share openly information about biodiversity,” says Neal Lane, science adviser to U.S. President Bill Clinton and vice-chair of last week's meeting.

    The new facility incorporates the recommendations of a bioinformatics working group, many of whose members are also leaders of existing efforts to compile and disseminate information about the range of species on Earth. The report lays out three major areas in which GBIF could have an impact.

    One involves compiling a definitive list of species names. “We need a unique way of referring to the elements of biodiversity. Most species have had more than one name and some dozens,” says Stephen Blackmore, keeper of botany at Britain's Natural History Museum, who served on the bioinformatics working group that proposed GBIF. To produce such a list, GBIF will work closely with Species 2000, an effort just under way to enumerate all known species of plants, animals, fungi, and microbes. “Its endorsement may also help us in obtaining additional funding,” says Frank Bisby of the U.K.'s Reading University, who chairs Species 2000.

    Another goal is to coordinate the development of new software to link databases that embrace the full range of biodiversity information including geographical, ecological, genetic, and molecular data. A third activity will be to digitize all biodiversity information, now usually embodied specimens in museums continents away from where the samples were collected. “Repatriation of data is a major impetus,” says Meredith Lane, vice president for biodiversity at the Academy of Natural Sciences in Philadelphia and a member of the bioinformatics working group. But that process, by which the host country would obtain electronic access to information stored in another country, will require an enormous and sustained effort. “We have 30 million insects on pins, many very small and fragile with tiny hand-written labels. At our current rate of progress, [cataloguing these specimens electronically] would take centuries,” says Blackmore.

    Of course, all this will take money. And despite the official go-ahead, none has yet materialized for GBIF. The working group has estimated that GBIF will end up coordinating some $40 million a year in ongoing work within member countries, and that GBIF itself can make an important contribution at an annual cost of $8 million a year. But such a budget, paid by member nations, is probably a few years away.

    As a first step, science ministers from Australia, Denmark, the United Kingdom, and the United States have signaled their intention to contribute toward the $2 million to $3 million needed to set up a six-person secretariat at a site to be determined. Australia and the United Kingdom are seen as likely bidders for the administrative headquarters, to open next year. Although the United States is unlikely to put in an application, says James Edwards of the National Science Foundation, it is strongly committed to the project. “There is some activity going on now to mobilize data, and there are sporadic efforts to put it on the Internet,” he says. “But there's no capacity for the one-stop shopping needed for nations to carry out the CBD and to develop their own biodiversity programs. That's what GBIF will do.”


    How to Get a Heart in The Right Place

    1. Gretchen Vogel

    CHARLOTTESVILLE, VIRGINIALike a child learning to put her hand over her heart for the Pledge of Allegiance, a developing embryo needs to know its right from its left. The heart goes on the left and the liver on the right, but how the embryo knows which is which is a long-standing puzzle. At the annual meeting of the Society for Developmental Biology here last month, one promising theory—that twirling “hairs” on embryonic cells set up the left-right distinction—gained strength.

    Scientists first proposed a connection between cilia—whiplike protrusions that can propel cells and help keep airways clear—and organ placement nearly 25 years ago. In 1976, Bjorn Afzelius described how human patients with a genetic defect called Kartagener's syndrome have immotile sperm and defective cilia in their airways—and about half have their organs on the wrong side (Science, 23 July 1976, p. 317). That connection led to speculation that cilia might somehow help to direct organ placement, but no one knew whether Kartagener's syndrome disables the cilia in the embryo as well.

    The old theory was resurrected 6 months ago, after cell biologist Nobutaka Hirokawa of the University of Tokyo and his colleagues reported that when they knocked out a gene involved in cilia assembly in mice, about half the animals had reversed left-right organ placement, and all lacked cilia on so-called node cells. These cells produce many of the signals that direct the early patterns in a mouse embryo, and the node is the site of some of the first molecular differences between left and right.

    When the team made microscopic videos of normal node cells, they found that their cilia rotated counterclockwise, unlike the back-and-forth motion of cilia on sperm or in airways. By tracking fluorescently labeled beads, the scientists determined that the cilia somehow swept the fluid surrounding the cells to the left. That might cause an as yet unknown signal to accumulate, eventually leading to asymmetric organ development. The lack of this lateral cue in the mutant strain could explain the 50% rate of organ reversal.

    But other researchers had trouble repeating the technically difficult observations, and many remained unconvinced. One concern was that the mice without cilia on their node cells might have other defects as well, so that something other than the cilia themselves could be the cause of the left-right disturbances. Even Yale University pediatric cardiologist Martina Brueckner, who had been working with a different strain of mutant mice that also suffer a 50% chance of left-right reversal, had her doubts. “It just seemed so weird,” she says.

    But, at the meeting, she announced that her team has taken a close look at the node cells in their mutant embryos, too. They found that these cells do have cilia, but they stand rigid and straight, instead of twirling. Without that motion, evidently, the signal drifts randomly left or right, which could explain the reversals.

    The observation boosts the theory that twirling cilia cause asymmetry, says cell biologist Chris Wright of Vanderbilt University. “Showing that they're rigid is tantalizing,” he says. But to really clinch the case, he says, someone needs to show that the cilia in yet another mutant mouse strain called inv, in which almost all the animals have reversed organs, twirl backward.


    EMF Researcher Made Up Data, ORI Says

    1. Dan Vergano*
    1. Dan Vergano writes for the Medical Tribune.

    In a blow to a research area hungry for credible findings, the federal Office of Research Integrity (ORI) reported last month that a biochemist “engaged in scientific misconduct … by intentionally falsifying and fabricating data and claims” in two studies on how electromagnetic fields (EMFs)—the kind shed by power lines and home appliances—affect living cells. The researcher, Robert P. Liburdy, formerly of the Lawrence Berkeley National Laboratory (LBNL) in California, has agreed to ask the journals to retract the results. “There's a lot of acrimony in the [EMF] debate, and this won't calm things down,” says Richard G. Stevens, a cancer epidemiologist at the Pacific Northwest National Laboratory in Hanford, Washington.

    Liburdy's findings were among the first to offer a plausible mechanism for a possible link between EMF exposure and cancer or other diseases. In a pair of 1992 papers of which he is the sole author, Liburdy offered evidence that EMFs increase the flow of calcium into lymphocytes, a kind of immune cell produced in the thymus. The papers created a stir, as calcium ions signal cells to turn genes on and off, and play a role in cell division. Because tumor growth is tied to cell proliferation, an alteration in calcium signaling could conceivably lead to cancer. But in an analysis obtained by Science, ORI states that “Liburdy's claims that EMF causes cellular effects related to calcium signaling [in three figures in the two journal articles] are not supported by the primary data.”

    Responding to an unknown whistle-blower's allegations of scientific misconduct by Liburdy, LBNL in January 1995 appointed a panel of four lab scientists to investigate. After reviewing raw data and interviewing Liburdy and other scientists, the panel concluded in a July 1995 report that Liburdy “deliberately created ‘artificial’ data where no such data existed” in a figure in FEBS Letters. In addition, it found, he fabricated data noise for a figure in the Annals of the New York Academy of Sciences “in order to mislead the reader.” These actions, the panel stated, “fall within the definition of scientific misconduct.” When contacted by Science, LBNL officials declined to comment, other than to confirm that Liburdy no longer works at the lab.

    Because Liburdy had been awarded more than $3.3 million in federal grants for his EMF research, ORI launched a formal review of LBNL's report in fall 1997. ORI approved a request by Liburdy for an interview with ORI staff and two outside experts, which took place in March 1998. At the meeting, Liburdy produced original data he had not shared with LBNL investigators, according to the ORI report. But the data failed to exculpate him: In its analysis, ORI accuses Liburdy of having lied to LBNL and ORI investigators, and it “concurs with [LBNL's] findings of scientific misconduct.” “Some of the numbers, essentially, he made up,” says John Krueger, an ORI investigator involved in the case.

    In a May 1999 agreement signed by Liburdy and ORI acting director Chris Pascal, Liburdy agreed to retract the tainted figures in the two papers and not to receive federal funds for 3 years. He “neither admits nor denies ORI's findings of scientific misconduct,” the document states. Liburdy did not respond to requests for an interview.

    The misconduct findings are unlikely to shift the playing field in EMF research. Since 1992, 20 to 30 scientific papers have looked at EMF exposures and calcium signaling, without settling the issue, says Christopher Portier, associate director of the environmental toxicology program at the National Institute of Environmental Health Sciences (NIEHS). In a report to Congress released on 15 June, NIEHS director Kenneth Olden states the scientific evidence that EMF exposures “pose any health risk is weak” and that mechanistic and toxicology studies “fail to demonstrate any consistent pattern.” The day before the report came out, National Institutes of Health officials had asked NIEHS to determine quickly whether any of Liburdy's research had influenced the report's conclusions, Portier says. The truth was simple, he says: “It had no impact whatsoever.”


    Frog Is a Prince of a New Model Organism

    1. Gretchen Vogel

    CHARLOTTESVILLE, VIRGINIAXenopus laevis, a fist-sized brown frog that is a favorite of developmental biologists, has an embarrassment of genes. For decades, biologists have studied its large, hardy embryos, transplanting bits of tissue to create monster tadpoles with two heads or missing tails—and in the process deciphering some of the key steps that shape a developing vertebrate. But in the age of molecular biology, X. laevis has a major drawback: Somewhere in its evolution, the frog's genome doubled, leaving the animals with four copies of most genes instead of the usual two. The extra genes make it nearly impossible for scientists to do the genetic studies that have been so powerful in flies, worms, and mice: interrupting the function of a gene and watching what goes wrong when it is missing.

    Now a frog from the damp floor of the West African rainforest, called Xenopus (Silurana) tropicalis, may provide the best of both worlds to developmental biologists, who crowded into a recent symposium* here to learn about it. The species is the only close relative of Xenopus that has a diploid genome, with just two copies of each gene, like people and most other vertebrates. It is smaller and easier to house than X. laevis and also becomes sexually mature in 4 or 5 months instead of 1 to 2 years, so scientists can quickly breed transgenic colonies.

    Yet the species are close enough that even minor steps in development are the same, and transcription promoters and other genes cloned from X. laevis seem to work just fine in X. tropicalis. The two species will be “two parts of the same system,” predicts cell biologist Marc Kirschner of Harvard Medical School in Boston. “All of this wonderful work and technology in laevis has been directly applicable to tropicalis,” he says.

    Kirschner was the first to import X. tropicalis to the United States. But most of the work on the new species has been in the laboratories of the symposium organizers: developmental biologist Enrique Amaya of the Wellcome/CRC Institute in Cambridge, U.K., and Robert Grainger of the University of Virginia in Charlottesville.

    Already, these labs are toying with the genetics of their new model. At the meeting, developmental biologist Lyle Zimmerman, a postdoctoral fellow in Grainger's lab, described some of the first transgenic X. tropicalis. A few years ago, Amaya and Kristen Kroll, now of Harvard Medical School, developed a technique for extracting nuclei from Xenopus sperm, treating the DNA so that it efficiently incorporates new genes, and then injecting the transgenic nuclei into eggs. A transgenic frog, with the new genes in all of its cells, then develops. Zimmerman and his colleagues have now used the technique to create frogs that express green fluorescent protein (gfp) in cells destined to become eye, heart, or the nervous system, allowing the scientists to observe the growing organs in live embryos. Although the gfp doesn't disrupt normal gene function, scientists can design DNA inserts that do interrupt key genes, then breed the frogs to produce offspring in which both gene copies are faulty.

    Such mutant frogs should prove a powerful tool for developmental biology. The ability to watch gene regulation without killing the embryo “is really unprecedented in a vertebrate,” says molecular biologist Barry Knox of the State University of New York, Syracuse. In mice, scientists can do even more sophisticated knock-out experiments, but they cannot observe the embryo as it grows inside its mother. And zebrafish, praised for their see-through embryos, are not as suitable for tissue transplant experiments as the larger frogs.

    To lay the groundwork for studying X. tropicalis, Grainger, Amaya, and a number of their colleagues hope to launch a major screen for mutant frogs, similar to the systematic screens done in flies and zebrafish. By causing random mutations and then watching their effects, scientists hope to tease out the genes that control various stages of development—and turn their frog into a prince of a model organism.

    • *“Moving into Xenopus tropicalis,” University of Virginia, 12 June.

  9. The Scientific Challenge of Hepatitis C

    1. Jon Cohen

    A virus that infects some 170 million people worldwide is causing rising rates of liver disease; like HIV, it is a wily foe for researchers developing drugs and vaccines. HIV may hold some lessons

    Nobody would have mistaken the international conference on hepatitis C, held last month at the National Institutes of Health (NIH), for an international conference on AIDS. For the past decade, international AIDS meetings have attracted roughly 10,000 researchers; only 775 scientists attended the 4-day hepatitis C meeting.* While drug companies working on anti-HIV drugs jam the exhibit halls at AIDS gatherings, not a single company set up a stand at the hepatitis C meeting. If activists infected with hepatitis C virus, or HCV, attended the gathering, none made their presence felt. Not a single press conference was held, reflecting the low media turnout (one). Yet HCV has infected an estimated 170 million people worldwide—more than four times as many as HIV—and, during the next few years, the number of annual U.S. deaths from HCV-caused liver damage and cancer may overtake deaths caused by AIDS.


    Despite these disparities, on a scientific level, the similarities between the HCV field today and HIV research in the 1980s are striking. “There's so much you can learn from HIV,” says David Thomas, a Johns Hopkins University clinician who studies and treats both viruses. Like their counterparts studying AIDS in the early 1980s, HCV researchers still can't grow the virus in laboratory cultures, and they don't know precisely how it infects a cell. They also have but foggy notions about the timeline between infection and illness, the so-called “natural history” of the disease. Currently available drugs, like early AIDS therapies, have serious toxicities and fail in most people—and no one knows for sure why some people respond to treatment and others do not. Nor have vaccines lived up to early hopes; just like HIV, HCV mutates rapidly, creating a swarm of different viruses in each infected person that can thwart antibodies easily. And, reminiscent of the struggles over patents on AIDS tests, lawyers from companies making diagnostics and drugs are firing salvos at each other over HCV patents (see sidebar, p. 28).

    HCV is not, of course, HIV. The hepatitis virus does not splice itself into the genes of a host, which means it may be easier to eradicate from a person's body. Indeed, some people become infected for several weeks and then naturally clear HCV from their bloodstream. HCV also does not target and destroy the immune system, and it may not cause clinical symptoms for decades in most of the people who become chronically infected. And, unlike HIV, HCV is rarely transmitted sexually; it seems to require direct blood-to-blood contact. Still, differences aside, HIV holds up an interesting mirror to the young HCV field, where “I don't know” remains the most common answer to a question.

    The NIH meeting came at an important juncture in the scientific battle against HCV. An improved therapy, approved just 6 months ago, has galvanized the field, and at least partial answers are now emerging to some of the most formidable unknowns about the virus. NIH also is increasing funding for HCV research (see table)—although the $33.6 million it plans to spend next year pales in comparison to the $1.8 billion to be spent on HIV. And as the barriers fall, more and more researchers—many of whom specialize in HIV—are being drawn to the field. Says Thomas: “It's like this smoldering fire that's finally starting to catch.”

    Elusive epidemic

    For years, clinicians knew that something in the blood supply was causing a small fraction of transfusion recipients to suffer short-lived flulike symptoms followed in some cases by liver disease years later. To distinguish the disease from better known forms of hepatitis, they referred to it by the ungainly but descriptive name of non-A, non-B hepatitis. Researchers from Chiron Corp. and the Centers for Disease Control and Prevention (CDC) finally unmasked the insidious agent in 1988. The next year, they jointly published papers in Science describing the new virus and a way to test for it in blood samples (21 April 1989, p. 359).

    To the surprise of many nonhepatologists, hepatitis C has little in common with its more famous cousins, A and B, except that all inflame the liver. HCV hails from a family known as Flaviviridae, and its close cousins include viruses that cause bovine diarrhea, hog cholera, and yellow fever. Carrying a single strand of RNA, HCV contains just one gene, coding for a polyprotein that is subsequently spliced into at least 10 functional proteins (see figure). Scientists have identified more than 100 strains of the virus and grouped them into six major “genotypes,” which tend to cluster in different regions of the world.

    No compelling clues point to where or when HCV first infected humans, and no other species appears to serve as a natural host to the virus. Yet studies clearly have shown that the main routes of transmission are by tainted blood transfusions and dirty needles used by injecting drug users, practitioners of folk medicine, and even public health campaigns (see sidebar, p. 27).

    Genetic blueprint.

    HCV's one gene produces a polyprotein that splits into at least 10 proteins.


    The development of a screening test in 1990 has virtually eliminated the spread of HCV through transfusions in industrial countries, and sharing contaminated needles is now by far the most common route of infection. As a result, CDC estimates that new U.S. HCV infections dropped from about 230,000 a year in the 1980s to fewer than 36,000 in 1996. But, because most of those infected in past decades are still alive, CDC estimates that perhaps 1.8% of the U.S. population harbors the virus. And as those patients age, HCV-related liver disease—which now accounts for 8000 to 10,000 annual deaths in the United States and is the single most common reason for liver transplants—likely will increase.

    Aside from direct blood contact, HCV is a very difficult agent to transmit. Even maternal-to-fetal transmission is low; no more than 6% of babies born to infected mothers will carry the virus. One question that is still being debated, however, is whether HCV can be transmitted sexually.

    At the meeting, the CDC's Miriam Alter reviewed several studies that “associate” HCV infection with having multiple sexual partners, leading to CDC's official conclusion that sex accounts for between 10% and 20% of the infection in the United States. “I think efficiency is very, very low, but it does happen,” Alter says. “Given that sex is frequent and 80% of the population have had more than one partner in a lifetime and that there are a number of chronically infected people, it makes sense.”

    Yet several lines of evidence argue against sexual transmission. As HIV proves, homosexual men in the United States transmit blood-borne viruses more efficiently than do heterosexuals, yet gay men have no higher HCV infection rates than do heterosexuals. And in a long-term study of 116 “discordant” couples, in which only one partner was initially infected with HCV, Harvey Alter (not related to the CDC's Miriam) and his co-workers at the NIH's Clinical Center found just 16 cases of new infections. In every one of those cases, the person who became infected had a history of injecting drug use or a blood transfusion. “So we don't have any direct evidence of sexual transmission,” concludes Alter. He now plans to compare the viruses in each of the 16 infected couples to confirm his suspicions that they did not infect each other.

    Who will become ill?

    Once scientists connect a bug to a disease, they can begin to unravel the natural history of a typical infection—how the disease progresses, and how long the process takes. Yet for HCV there appears to be no such thing as a typical infection. The severity of the disease varies greatly from person to person and—to the frustration of clinicians and patients—there are few reliable indicators to predict who will do well or badly.

    Evidence accumulated over the past few years indicates that the immune systems of 15% to 25% of people infected with HCV will overcome the virus during the initial infection and clear it from the bloodstream. The remaining 75% to 85% will develop a chronic infection. HCV targets liver cells, called hepatocytes. As hepatocytes die off, fibrous tissue forms, which scars the liver, preventing blood from passing through it and leading to the life-threatening condition known as cirrhosis; this occurs in perhaps 10% to 20% of chronically infected people. Another 1% to 5% of the chronically infected also develop a liver cancer called hepatocellular carcinoma. Yet, as several studies presented at the international meeting show, the majority of patients have none of these symptoms even 20 years after infection.

    NIH's Alter and Jay Hoofnagle from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) reported results from a 7-year study of more than 400 would-be blood donors who had tested positive for HCV and whose infection could, in most cases, be traced to a transfusion or injection. Even though they had been infected for an average of nearly 20 years, only 13% had severe fibrosis and a mere 2% had cirrhosis. These results closely match those from an Irish study published in the 22 April New England Journal of Medicine that charted disease progression over 17 years in 376 women who had received contaminated blood products in the 1970s. And a study by NIDDK's Leonard Seeff looked at 8568 blood samples stored by the U.S. Air Force between 1948 and 1954 and found that 17 tested positive for HCV antibodies; current records revealed that only one of those infected individuals (5.8%) died from liver disease.

    Alter finds these results reassuring for the majority of patients. “What's needed in this field is some perspective,” he says. “Hepatitis C is getting such big play, mostly with a sense of alarm. We spend a lot of time trying to calm people down. The real issue is, What proportion of people are going to reach the bad outcome?”

    For individual patients, however, the real issue is, “Am I going to have a bad outcome?” So far, clinicians are hard put to provide an answer. They have found little correlation, for example, between the amount of virus in a patient's blood—the “viral load”—and disease progression. And several studies have found that none of the six HCV genotypes appears to be more pathogenic than the others. Alcohol consumption may increase the risk of disease progression—two of the seven patients with cirrhosis in the Irish study were heavy drinkers, for example—but, again, the risks appear to vary. Moreover, data from NIH's blood donor cohort presented at the meeting indicate that one commonly used test to assess liver damage may have little predictive value, too. This blood test measures an enzyme called alanine aminotransferase (ALT).

    Hepatocytes release ALT when they die, so ALT levels should provide an indirect measure of how much damage HCV is doing to a liver. But in a study that began in 1993, Marc Ghany, a researcher in Hoofnagle's NIDDK lab, separated 60 HCV-infected people into three groups based on whether they had normal, mildly elevated, or moderately elevated ALT levels. Over the next 5 years, most participants' ALT levels remained in the same category. But liver biopsies showed that people with normal ALT levels had, on average, worsening of their fibrosis, while those in the mild and moderate categories had slight improvements. Hoofnagle cautions that these preliminary data come from a relatively healthy infected group, but he says they are “disturbing” nonetheless. “If you follow a person and ALTs are normal, we always thought that their [degree of fibrosis] would be near normal,” says Hoofnagle. “This finding goes contrary to everything we believe.”

    Researchers say a better understanding of the natural history of HCV infections will require a major, long-term cohort study, such as the 15-year Multicenter AIDS Cohort Study (MACS) that has shed valuable light on the natural history of HIV. Leslye Johnson, who heads the enteric and hepatic disease section at the National Institute of Allergy and Infectious Diseases (NIAID), notes that a 1997 “strategic plan” put together by her institute called for such studies, but they have not been funded. “It's a money issue,” says Johnson. Indeed, Johnson, whose institute this year spent about $7.7 million on HCV, says lack of funding is crimping other areas as well. “I put together an initiative list last year for internal consumption,” she says. “I said I'm not going to ask for what I know I can get, I'm putting down what I need.” Johnson's figure: $40 million. “Branch chiefs looked at it and rolled their eyes.”

    A cultural barrier

    It's not money, however, that hepatitis C researchers mention when asked what the field needs most. As Frank Chisari, a leading hepatitis immunologist at The Scripps Research Institute in La Jolla, California, puts it, “We desperately need a culture system.” To date, no one has been able to grow HCV reliably in a laboratory culture of cells, a lack that has slowed critical studies of everything from drugs to vaccines to basic understanding of the viral life cycle. Little wonder, then, that the showstopper at the NIH conference was a report of a new HCV culture system, which is described on page 110 of this issue.

    Developed after 5 years of effort by Ralf Bartenschlager and colleagues at the Johannes-Gutenberg University in Mainz, Germany, the culture system does not actually grow HCV itself. Rather, Bartenschlager's group engineered a stretch of DNA that contains the mirror image of a portion of HCV's RNA. Bartenschlager injected this “replicon,” which codes for HCV's nonstructural proteins but not its core or surface proteins, into immortalized human cell lines. The replicon then copied itself to high levels, which he showed both by polymerase chain reaction assays and by measuring viral proteins.

    “It's a groundbreaking study,” says NIDDK's Jake Liang, who with Hoofnagle co-organized the conference. “People have to be cautioned—this is not productive infection. It does not generate virus. Still, it's a major step in the right direction.” Stanley Lemon of the University of Texas Medical Branch at Galveston adds that “if these results hold up, they'll be enormously useful for drug screens.”

    Because the replicon does not produce whole viruses and their attendant envelope proteins, however, researchers cannot use it to determine how HCV infects cells—a critical question that has been frustratingly difficult to answer. As a team led by Sergio Abrignani of Chiron's Siena, Italy, branch reported in the 30 October 1998 Science (p. 938), one of HCV's surface proteins binds to a cell surface receptor called CD81. But the group did not show that HCV used the receptor to infect cells, and many researchers suspect that it is only part of the story. “CD81 is very intriguing, but no one has proven that it is needed for entry,” says Charles Rice, a prominent HCV molecular biologist at Washington University School of Medicine in St. Louis, Missouri.

    Treatment: Limited success

    Until last year, people infected with HCV had only one choice of treatment: three injections a week of interferon—a chemical messenger naturally produced by the immune system—for up to a year. For most, it wasn't much of a choice; the drug eliminated the virus in less than 20% of patients. But last fall, the field was jolted by the results of two major trials that showed that combining interferon with an antiviral drug called ribavirin at least doubles the chances of success in people who had never been treated.

    View this table:

    The first of the two studies, published in the 31 October 1998 Lancet, compared 48 weeks of treatment with either interferon a-2b alone or the two-drug combination in 832 HCV-infected people. Forty-three percent of the patients on the combination therapy had no detectable HCV RNA in their blood 24 weeks after treatment, while only 19% of those taking interferon alone had a similar “sustained” virological response. One promising sign: Unlike HIV, HCV does not appear to lurk in hard-to-treat reservoirs from which it will reestablish an infection.

    The second study, published in the 19 November 1998 New England Journal of Medicine, involved 912 patients. Again, 24 weeks after treatment stopped, 38% of those who took both drugs had undetectable HCV RNA, compared to 13% of those on monotherapy. Just 1 month after the study was published, the Food and Drug Administration approved the combination therapy for previously untreated patients.

    Promising as these results are, the drugs are expensive—a 48-week treatment costs nearly $20,000—and cause debilitating flulike symptoms in most people, leading about 20% of the patients in both trials to stop treatment early. “Interferon is a difficult drug to take, and ribavirin makes it worse,” says NIH's Alter. More sobering still, the drugs are less effective against the most common strain in the United States, known as genotype 1, which is responsible for about 70% of U.S. infections. The New England Journal study, for example, found that only 28% of people with genotype 1 had a sustained response, compared to 66% of those infected with other genotypes. (Other factors also affect the likelihood of success: Women, patients under 40, those with low viral loads, and those without severe liver damage all tend to do better.)

    These results have sparked a sharp debate over whether to treat asymptomatic patients. Alter says the high failure rate and the toxicity make him cautious. “Drug companies are pushing everyone to treat everyone, but often it's a mild disease,” he says. And he notes that the treatment seriously disrupts most infected people's lives. “These people all feel well and all of a sudden they feel badly,” he says, echoing a refrain that is often heard when asymptomatic HIV-infected people begin taking AIDS drugs.

    Why some people fail therapy and others succeed is a mystery, in part because nobody knows precisely how interferon and ribavirin work. But a study published on page 107 of this issue offers one possible explanation for why some HCV strains are more resistant to interferon than others.

    Michael Lai of the University of Southern California in Los Angeles and his co-workers focus on an enzyme called protein kinase PKR that derails viruses by inhibiting their protein synthesis. Interferon turns up the rate at which cells make PKR. But the researchers' data suggest that HCV has a weapon against the enzyme: one of its two surface proteins, dubbed E2, which inhibits the activity of PKR. Based on sequence analyses of E2, they further contend that the protein made by HCV genotype 1 is especially good at binding to PKR and blocking its function. “This is an interesting observation that has to be further explored,” says Michael Katze of the University of Washington, Seattle, whose lab focuses on a nonstructural HCV protein called NS5A that inhibits PKR. “There's no precedent for an envelope protein being a PKR-like molecule.”

    Clinicians now are tweaking treatment regimens and testing improved versions of both interferon and ribavirin. But, as with anti-HIV drugs, bigger gains are expected from compounds tailored specifically to attack key HCV proteins. Several companies are searching for inhibitors to enzymes that HCV uses to copy itself: protease, helicase, polymerase, and replicase.

    The most popular target to date has been the “serine” protease, one of two viral enzymes that helps clip HCV's polyprotein into functional proteins. “Everybody picked the protease target first simply because of the success with HIV [protease inhibitors],” says Lewis “Rusty” Williams, Chiron's chief scientific officer. Like other companies, Williams says Chiron—which has teamed up with Pharmacia & Upjohn to develop anti-HCV drugs—is pursuing other targets, too. “We have progress in several of those targets, and it's hard to say which is furthest,” says Williams. “I'm sure a number of companies are at work on the same targets. It's a race.” Neither Chiron nor its competitors, however, have indicated that they are far enough along to set a timeline for human trials. And the dearth of presentations on these efforts at the meeting indicates that the companies are holding their cards closely to their chests.

    Vaccine vacuum

    While companies are elbow-to-elbow in the HCV drug development race, Chiron—in part because of its patent position (see sidebar, p. 28)—has had the vaccine field largely to itself. It is facing a tough challenge. Like HIV vaccine developers, it must contend with the fact that antibodies directed against rapidly changing viral proteins are unlikely by themselves to offer protection. In one test, for example, Chiron found that chimps vaccinated with genetically engineered versions of the virus's two surface proteins, E1 and E2, failed to fend off infection when they were inoculated with a strain of HCV that differed from the strain used to produce the vaccine. This has led the company to put increasing emphasis on the cell-mediated arm of the immune system, which uses killer cells and other strategies to clear the body of already infected cells.

    Drug targets.

    Protease enzyme (left) and helicase unwinding RNA (right)


    In the April issue of Immunity, Stanford University immunologist Stewart Cooper, in collaboration with Chiron researchers, published the results of a study with chimpanzees that hints at the power of this approach. The researchers inoculated one chimp with E1/E2 antibodies derived from humans and another with the Chiron E1/E2 vaccine. The study also included two other animals (LouLou and Todd) that had previously been injected with HCV but had cleared the virus, and two naïve animals as controls. When the animals were “challenged” with HCV, the first two and the controls readily became infected. LouLou and Todd, however, resisted HCV. Subsequent analyses revealed that while neither LouLou nor Todd had antibodies from their previous exposure to HCV, both had robust killer cell responses. This implies that cell-mediated immune responses may have wiped out their earlier infections and could now protect them from subsequent infections.

    To trigger a strong cell-mediated immune response, Chiron is investigating vaccine strategies that produce viral proteins inside the body's cells. One such approach involves a so-called DNA vaccine, in which the injected preparation contains viral DNA by itself. Chiron also is exploring the possibility of stitching viral genes into safe viruses that then can infect cells and cause a mock infection.

    Chiron may soon have some competition in these efforts. Scripps's Chisari and Chris Walker—a co-author on the Immunity paper who recently left Chiron for Ohio State University in Columbus—are working with a San Diego biotech, Epimmune, to develop an HCV vaccine that exploits cell-mediated immunity. And on the academic front, Fred Prince at New York Blood Bank is combining an HCV DNA vaccine with one that stitches HCV genetic material into a fowlpox virus; and NIAID's Robert Purcell is also exploring the DNA vaccine approach.

    Walker, a former AIDS researcher, cautions that even if these efforts produce a promising vaccine, testing it in humans will be difficult. In the United States, efficacy tests probably could be done only in injecting drug users, a difficult group to follow for the length of a trial. Tests in poor countries would run into the same logistical and ethical problems now facing trials of anti-HIV vaccines. “I think we have a lot to learn from HIV,” says Walker. “The efforts going on there are going to blaze the trail for HCV.”

    For most of the HCV-infected world, the development of vaccines will be crucial. Although hepatitis C could become a curable disease in a decade or so, if the next generation of drug therapies live up to their promise, the treatments will only help people in those countries that can afford the drugs. And that is perhaps the most sobering lesson from AIDS.

    • *Sixth International Symposium on Hepatitis C and Related Viruses, 6–9 June, National Institutes of Health, Bethesda, Maryland.

  10. Public Health Effort Unwittingly Spread HCV

    1. Jon Cohen

    Roughly 24% of the people in Egypt are estimated to carry hepatitis C virus (HCV), making it the hardest hit country in the world. Researchers have long suspected that the culprit might be a decades-old strategy to combat a parasitic disease known as schistosomiasis. Now a study headed by Thomas Strickland, an epidemiologist at the University of Maryland, Baltimore, takes this idea from the realm of the possible to the probable. Maryland's Christina Frank, who presented the data at the Sixth International Symposium on Hepatitis C and Related Viruses for Strickland's lab and their collaborators at Ain Shams University in Cairo, Egypt, says the schisto-HCV link “may very well be the world's largest iatrogenic transmission scenario known to date.”

    Schistosomiasis in Egypt dates back at least to the time of the Pharaohs. Caused by a flatworm that propagates in water snails, the disease attacks the intestines, bladder, liver, and other organs. In 1918, Frank explained, physicians began popularizing various injectable treatments, called parenteral antischistosomal therapy, or PAT, which typically required 10 to 12 injections and was usually given with reusable syringes. The campaigns began tapering off in the 1970s when oral schisto drugs became available.

    By studying schistosomiasis archives at the World Health Organization, PAT records at the Egyptian Ministry of Health, census records, and HCV infection surveys of 10,000 Egyptians, the researchers pieced together the link between the two diseases. Age and demographic data indicate that people who were most exposed to PAT had the highest HCV prevalence. “That's the first time the data convincingly show this,” says Robert Purcell, a virologist at the U.S. National Institute of Allergy and Infectious Diseases, whose laboratory has played central roles in the study of every hepatitis-causing virus.

  11. Chiron Stakes Out Its Territory

    1. Jon Cohen

    When Chiron Corp. first considered joining the search for the agent that causes non-A, non-B hepatitis in the 1980s, the Emeryville, California, biotech turned to one of its scientific advisers, Harold Varmus, for his opinion. Varmus, who later became director of the National Institutes of Health, suggested that the company place its bets elsewhere because “nobody has been able to crack this problem,” recalls Robert Blackburn, Chiron's chief patent attorney. But Chiron went ahead anyway, and in May 1988 it announced the discovery of what is now known as hepatitis C virus, or HCV. Blackburn tells this story to explain why Chiron is aggressively defending its HCV-related patents. The company took a huge risk in pursuing this research, he says, and “that investment would not have been made if there wasn't a patent system. This shows the patent system is working and working right.”

    Chiron's patents cover HCV protease.STRUCTURE: NANHUA YAO AND PATRICIA WEBER/SPRI

    Many of Chiron's competitors would beg to differ. Their activities have drawn a barrage of Chiron-initiated patent infringement lawsuits, which some researchers say are having a chilling effect on the field. Chiron has filed suit in Europe, Australia, and the United States against Murex, Organon Teknika, and Hoffmann-La Roche, charging the companies with selling HCV blood tests without paying licensing fees. Last July, it went after four companies involved in HCV drug research. Chiron also has refused to license its technologies to companies interested in developing HCV vaccines, although Chiron officials say they have recently decided to loosen the strings.

    One of the highest stakes suits targets Hoffmann-La Roche and its subsidiaries. Roche markets HCV detection tests that rely on the ultrasensitive polymerase chain reaction (PCR) assay. Tom White, senior vice president of R&D at Roche Molecular Systems, says its PCR test is the “gold standard” for HCV blood screening. “If Chiron were to block this, the impact to the medical community would be highly negative,” he asserts.

    Roche argues that its test is covered by the rights to PCR that it acquired from Cetus, a company that Chiron subsequently purchased. “Chiron is obligated to uphold Cetus's agreement,” says White, who notes that Roche negotiated for “broad” rights. Blackburn says he is “amazed” by this claim, which he calls “both desperate and laughable, and we expect it to be thrown out of court.”

    Roche has a second line of argument, however, colloquially known as the “Daniel Bradley” defense. Bradley collaborated with Chiron on non-A, non-B hepatitis when he worked at the Centers for Disease Control and Prevention (CDC) in Atlanta, and he has long claimed that he played an instrumental role in the discovery of HCV. U.S. government attorneys agreed, so in 1990, Chiron reached a settlement to avoid litigation: In exchange for U.S. patent rights, the company would pay Bradley $337,500 and the government $2.25 million. In 1994, however, Bradley sued Chiron, arguing that he should have been named an inventor on the patent and only signed the settlement agreement for health reasons (Science, 6 January 1995, p. 23).

    Bradley lost the suit and his appeals, but in the meantime, he assigned to Roche his non-U.S. “rights” to HCV technology. Roche attorney Melinda Griffith explains that Bradley never waived those rights, and she further contends that he's an inventor, which “raises questions about the patent's validity.” But Blackburn says CDC reached the 1990 settlement because its attorneys concluded that Bradley had not played an inventorship role. “That's totally bunk,” retorts Brent English, Bradley's attorney. Roche and Bradley now have filed counterclaims against Chiron.

    The lawsuits involving HCV drug R&D center on efforts to find drugs that block the viral protease enzyme, on which Chiron holds patents. The company, arguing that its competitors need this enzyme to screen for compounds that inhibit it, filed suit against Agouron, Gilead, and collaborators Vertex and Eli Lilly to try to force them to pay licensing fees and then royalties if one of their protease inhibitors goes to market. “This is reaching fairly deep into the R&D process,” complains Charles Rice, an HCV molecular biologist at Washington University School of Medicine in St. Louis, who worries that such claims are scaring away would-be HCV drug developers. Indeed, in documents filed in court, Agouron charges Chiron with “unlawful monopolization” of the U.S. HCV protease inhibitor market and says its R&D “has been delayed and impeded.”

    Chiron's chief scientific officer, Lewis “Rusty” Williams, counters that his company has licensed its HCV protease technology to a half-dozen others. “There's no impediment to getting involved with this,” says Williams. “The downstream payment is quite small relative to the total revenue that such a product could generate.”

    In its legal response to Chiron's suit, Vertex invoked a defense that many patent lawyers believe is ripe for a test case: safe harbor. The U.S. Congress in 1984 gave drug companies a “safe harbor” exemption if, in pursuit of data for the Food and Drug Administration, they tested a generic version of a patented drug before the patent expired. Does this exemption apply to all biomedical R&D? “That's one of the great unsolved mysteries of contemporary patent metaphysics, the reach of the research exemption,” says patent law expert Robert Merges of the University of California, Berkeley.

    The safe harbor defense outrages Blackburn. “Legally it's poppycock, and morally it would be a tragedy,” he says. “If you accept that argument, then the technology has no value, because it would be impossible to infringe.”


    A New Look Into Neandertals' Noses

    1. Constance Holden

    Noses hold clues to how these ice age humans lived and breathed. To read those clues, researchers have to rely on models and extrapolation

    Neandertals, that unique breed of hominids who frequented the forests and caves of Europe for perhaps 200,000 years, have been an object of intense interest since the first one was discovered in Germany's Neander Valley in 1856. They have also been a focus of controversy. Whether or not Neandertals and early modern humans ever clashed in war, as some researchers have suspected, anthropologists have been engaging in their own “Neandertal wars” over whether these heavyset bipeds were basically like us or built differently enough to qualify as a separate species.

    All sides agree that Neandertals tended to have prominent brow ridges, heavily muscled bodies—and big noses. Most of the external part of the nose is made from cartilage, which doesn't last. But from the shape of the nose hole, researchers conclude that the typical Neandertal schnoz was high, wide, and projecting. Crucial internal structures called turbinates, paper-thin, scroll-like bones that lie along the insides of mammalian noses, have not survived in fossils. But because these humans evolved in northern climes—actual glacial conditions in the case of later European Neandertals—some scientists say their turbinates would likely have been large to increase the mucous membrane-covered internal surface area available to warm and humidify cold, arid air.

    Scientists are now engaged in a flurry of research, measuring and modeling noses to try to push beyond these generalizations and see just how Neandertal noses differed from those of Homo sapiens both modern and archaic. And lately, as a half-dozen presentations at the meetings of the Paleoanthropology Society and the American Association of Physical Anthropology this spring in Columbus, Ohio, showed, studies of the Neandertal nose and face are allowing researchers to look beyond the species controversy. Led by the Neandertal nose, anthropologists hope to learn more about Neandertal breathing and energy use, and thus about the hard-driving Neandertal lifestyle, says anthropologist Jeffrey Laitman of Mount Sinai Medical Center in New York. Although some scientists continue to probe the species question, he says, others are “trying to get at the real issues and look beyond whether it's Homo lumpus or Homo bumpus.”

    The flurry of interest was kicked off in 1996 in the Proceedings of the National Academy of Sciences, where evolutionary biologist Jeffrey H. Schwartz of the University of Pittsburgh and anthropologist Ian Tattersall of the American Museum of Natural History in New York claimed to have discovered several nasal features unique to Neandertals. The most important finding, gained from examinations of skull bones of five European Neandertal specimens, was that they, unlike modern or ancient Homo sapiens, had distinctive bumps of bone—separate from their turbinates—on each side of their nasal apertures. The authors called these “medial projections” and suggested they might have served to increase surface area within the nose, helping to warm and moisturize inhaled air.

    Schwartz and Tattersall's paper set off a flurry of work on those bumps, and the results have been contradictory. Some anthropologists have failed to find the bumps on other Neandertal specimens, while others claim to have found them on modern human skulls as well. At the meeting, anthropologist Robert Franciscus of the University of Iowa, Iowa City, reported that he tried to settle the issue by making more than 40 measurements from each of 523 modern skulls as well as those of 200 fossil hominids, including all available Neandertals. Neither Neandertals nor moderns had the kind of bumps Schwartz and Tattersall had described, he says. Rather, every population had some sort of simple bump that probably represents the “roots” of long-gone turbinates.

    Schwartz and Tattersall, however, say their own analysis of additional specimens bears out that Neandertal nose bumps are clearly different from anything found in other fossil humans. But even though the battle of the bumps is unresolved, it has sent specialists flying back to their specimens for more study, says anthropologist Steven Churchill of Duke University, looking for new insights into how Neandertals breathed.

    Churchill himself recently decided that one way to find out more about the Neandertal nose was to model noses in action. He says airflow turbulence makes the nose more efficient at heating and humidifying the air stream, so Neandertal noses would be expected to promote turbulence. But, Churchill says, “unlike modern humans living in cold and dry environments, Neandertals also had wide nasal passageways,” more like Africans—which facilitates smooth, or laminar, airflow with minimal heating.

    To understand such an unusual combination of traits, Churchill made clear acrylic casts of human noses, from molds taken from 10 medical school cadavers. He then put tubes up the casts, which represented a range of European nose types, and siphoned dyed water through them to study the flow dynamics. “We tested the argument that a number of nasal features [including the large aperture and protruding turbinates] that are accentuated in Neandertal noses function to induce turbulence,” he says. What he found, however, was that only one feature—downward-directed nostrils, found in modern Europeans and, he believes, Neandertals—increased turbulence. Large turbinates actually seemed to reduce it, says Churchill.

    He suggests that the wide noses and, to an extent, the large turbinates were a trade-off. Stocky Neandertals, like heavy-duty trucks, were very heavy energy users, and simply breathing in enough oxygen would have required as much as 1000 calories a day—double the amount a modern male needs. The Neandertal nose shape may have made it less efficient at heating, but it also reduced the drag that limits airflow in a narrow nose, he speculates.

    Other researchers praise Churchill for rolling up his shirt-sleeves and designing an experiment to address these issues. But some say he is making unwarranted assumptions. Anthropologist Patrick Gannon, who directs the Sinus Research Laboratory at Mount Sinai School of Medicine in New York City, says there is no way of knowing the size of Neandertal turbinates or whether their nostrils were downward-directed. Gannon also says that no matter what the shape of the nose, the airflow is normally laminar when humans inhale, and he thinks it unlikely that Neandertal breathing would be any different. Franciscus, too, says Churchill makes too many assumptions about the size of Neandertal noses. “They're no different from cold-adapted early modern humans,” he says. Churchill responds that turbinate remains on two Neandertal specimens indicate they were large, and the large base of the Neandertal septum indicates a downward turning nose.

    As the nose exchanges continue, the research is also spilling over into other areas of the Neandertal head. For example, Franciscus says his measurements do confirm one “completely unique” Neandertal feature: the shallow depth of the throat. The vertical distance between the back of the roof of the mouth (the hard palate, which is also the floor of the nose) and the hole where the spinal cord exits the skull is much shorter in Neandertals than in either living or early modern humans, a characteristic first noted by Laitman more than 20 years ago.

    At the time, Laitman thought this trait might have limited Neandertal speech, but scientists are now trying to focus on less speculative issues. Franciscus says he sees the short throat as a by-product of other Neandertal traits. He notes that they, like modern humans, had big brains. But they were more like the earlier Homo erectus in their patterns of facial growth. To fit this big brain on top of their primitive face, they had to alter their braincase—and ended up with an unusually short throat, he theorizes.

    Laitman's group thinks that Neandertal uniqueness also extends to respiratory tracts, inner ears, eustachian tubes, and sinuses. “I say when you take the upper respiratory tract together with [these other features] we may be looking at a very distinctive bauplan,” he says. The bottom line, he contends, is that all these features support the notion that Neandertals relied more heavily on nose breathing than do modern humans. Franciscus, who thinks Neandertal noses were nothing special, isn't persuaded, insisting that “from everything we can measure about their internal nasal morphology, they breathed the same as we do.”

    With plenty of disagreement left, it's probably going to be a long time before scientists reach consensus on how Neandertals breathed—let alone what that might say about their relation to ourselves. But paleontologist Fred Spoor of University College London believes the field is taking a promising direction. There's less “storytelling” going on, as research shifts from the species question to how Neandertal physiology worked. “There's a bit of a new school of people saying … let's try to make a testable hypothesis,” he says, and that's “ultimately a more scientific approach.”


    Potential Target Found for Antimetastasis Drugs

    1. Elizabeth Finkel*
    1. Elizabeth Finkel writes from Melbourne, Australia.

    Researchers have finally cloned the gene for the enzyme heparanase, which helps cancer cells escape to new sites in the body

    Cancer cells are dangerous not so much because they've lost the brakes on their growth. Rather, it's their ability to metastasize—escape from the original tumors and spread through the circulation to new sites in the body—that makes cancer so tenacious and deadly. Now, researchers have gotten their hands on a key enzyme that helps cancer cells roam the body and may thus be a good target for anticancer drugs.

    Like the patrolling cells of the immune system, spreading cancer cells have to be able to breach such barriers as the extracellular matrix (ECM), the glue that holds cells together in tissues, and the basement membranes that surround the blood vessels. These consist of proteins embedded in a fiber meshwork consisting mostly of a complex carbohydrate called heparan sulfate. In previous work, researchers had cloned several genes for the enzymes, called proteases, that cancer cells use to break down the protein portion of the ECM and basement membranes.

    But even though researchers suspected for nearly 15 years that the cells also produce an enzyme that snips the heparan sulfate meshwork, that enzyme had eluded them—until now. Two groups, one led by Christopher Parish of the John Curtin School of Medical Research (JCSMR) in Canberra, Australia, in partnership with Progen Industries in Brisbane, and the other by Israel Vlodavsky at Hadassah-Hebrew University in Jerusalem and Iris Pecker of the biotech firm InSight Ltd. in Rehovot, Israel, report in the July issue of Nature Medicine that they've finally cloned the long-sought heparanase gene.

    The wait was apparently worth it. Although metastasizing cancer cells may produce as many as 15 different matrix-digesting proteases, the new work suggests that there is only one heparanase. Thus, if its activity can be inhibited—and indications are that it can be—other heparanases shouldn't be around to cover for it. “This is very exciting and surprising,” says Lance Liotta, a metastasis expert at the National Cancer Institute in Bethesda, Maryland, whose own work has focused on the proteases.

    What's more, a blow to heparanase apparently packs a double punch. Besides inhibiting cancer cells' ability to roam, blocking heparanase also hinders the formation of the new blood vessels that feed tumors, perhaps because the enzyme helps the vessels' growing tips penetrate new tissue.

    Researcher first made the connection between metastasis and heparanase in the mid-1980s. Three groups—Garth Nicolson's at M. D. Anderson Cancer Center in Houston as well as Parish's and Vlodavsky's—were following up on the finding that the natural anticoagulant, heparin, inhibits the spread of cancer in animals. The prevailing belief was that heparin worked because it prevented platelets from clotting around cancer cells, an event likely to help the cells lodge into, and ultimately penetrate, the vessel wall. But “heparin” is a family of molecules, only some of which inhibit clot formation. And the three groups independently showed that it still inhibits metastasis, even when depleted of its anticlotting activity.

    The researchers then traced that effect to molecules that inhibit heparanase, an enzyme then known only on the basis of its ability to break down heparan sulfate. Both Parish and Vlodavsky had already shown that heparanase helps immune cells traverse blood vessel walls on their way to infection sites. The evidence that it might be doing something similar for cancer cells, says Parish, “immediately made people think about getting better [heparanase] inhibitors.”

    But getting the pure enzyme that researchers wanted for their studies proved to be difficult. Not only is heparanase unstable, but the only assay then available was slow and cumbersome, which often meant the enzyme died before it could be recovered. Indeed, along the way, different laboratories appeared to be chasing different enzymes ranging in size from 8 to 137 kilodaltons.

    Nevertheless, the Israeli group finally managed to purify heparanase from a human liver cancer cell line and also from human placenta, while the JCSMR group purified it from human platelets. After determining partial amino acid sequences of the purified proteins, the researchers then screened databases looking for gene sequences that could encode those amino acid sequences. And contrary to expectations that there might be more than one heparanase, both groups found themselves with the same gene—the only one like it in the databases.

    Experiments by the two groups confirm that the gene they have cloned aids the spread of cancer cells. When Vlodavsky and his colleagues introduced a copy into nonmetastatic mouse melanoma and lymphoma cancer cells, they turned into rampantly malignant cells that colonized the lung and liver when injected into mice. And Parish, looking at several different types of rat cancer cells, found that their invasiveness correlates with the activity of their heparanase gene.

    Conversely, inhibiting the enzyme inhibits cancer metastasis. In work in press in Cancer Research, Parish reports that a previously identified inhibitor of heparanase called PI-88 decreased by 90% the number of lung tumors formed by breast cancer cells injected into rats. It also cut the blood supply of the primary tumors by some 30% and—perhaps as a result—slowed their growth by half. The encouraging animal results have already led Progen to test the safety of the inhibitor in healthy volunteers. “The drug was well tolerated over the few days” it was tested, says Parish.

    A trial in cancer patients should begin soon in Australia, and it is unlikely to be the last. InSight has its own active program to look for heparanase inhibitors, and other companies may well follow suit. “Now that the sequence is published, the competition [to find heparanase inhibitors] will be tough,” Vlodavsky says.


    Java Applet Lets Readers Bite Into Research

    1. James Glanz

    Luis Mendoza first began writing scientific demonstrations in the computer language called Java for an introductory astronomy course at the University of Washington (UW), Seattle. The language had several advantages for teaching students about such hard-to-visualize concepts like redshift and parallax: It is interactive, it works the same way on every type of computer, and its “applets,” or programs, run easily on browsers linked to the World Wide Web. Now Mendoza, a UW graduate student, has gone far beyond Astronomy 101 with what may be the first electronically published astrophysics paper to use an interactive Java applet.

    Although the project was modest, allowing users to run a calculation of element-forming processes during the big bang, it bears watching in a field that has been a scientific bellwether in the use of electronic media and the Web. “'Pioneer' sounds a little grand,” says Craig Hogan, the UW astrophysicist who recruited Mendoza for the project, “but it does make the precise predictions of the theory more accessible.” Subir Sarkar, a physicist at the University of Oxford in the United Kingdom, sees it as “a handy tool to enable observers to interpret their data.”

    Java, developed and copyrighted by Sun Microsystems of Palo Alto, California, is already in wide use for business and consumer applications, especially on the Web, where recent versions of most browsers can run Java applets. Java has also made inroads into teaching. Mendoza, for example, had developed an Astronomy 101 site containing applets such as a three-dimensional simulation of parallax, showing a star field and Earth moving in its orbit, seen from an angle chosen by the viewer.

    “There are all these young people in Seattle who do all this groovy stuff,” says Hogan. Taking advantage of that milieu, Hogan asked Mendoza to write an applet incorporating recent calculations by Sarkar and others on the generation of the light isotopes deuterium, helium-3, helium-4, and lithium-7 in the big bang. The amounts created depend on the overall mass density of the universe. The applet lets the user specify the value for one of the isotopes and then shows—within error bars reflecting imperfectly known nuclear reaction rates—the predicted values for the others and for the overall mass density.

    The applet can be found on the Los Alamos preprint server and at Researchers like Sarkar are already asking for upgrades, and at least one other astrophysicist—Paul Steinhardt of Princeton University—wants to take things a step further. He's planning to incorporate Java into a more comprehensive set of cosmology calculations recently published in Science (28 May, p. 1481) and nicknamed “the cosmic triangle.” Whether for research or for Astronomy 101, says Mendoza, the goal is the same: “It's just trying to communicate or teach what you have found.”

  15. AFRICA

    Danes Bring DNA Analysis To the Heart of Africa

    1. Lone Frank*
    1. Lone Frank is a writer in Copenhagen, Denmark.

    African biologists have long struggled to study their wealth of wildlife without DNA tools. Now funding from Denmark has created a state-of-the-art lab

    KAMPALA, UGANDAPopulation geneticist Peter Arctander has long provided an unofficial DNA analysis service for colleagues in Africa. Researchers, mostly from Uganda where Arctander has strong links, would ship their samples to his lab at the University of Copenhagen and, if the tissue reached Denmark in good condition—not always the case—the analysis would come back several weeks later by mail. Arctander has been providing this service because many African researchers have had nowhere else to turn: Not a single university lab in sub-Saharan Africa, outside South Africa, has had modern DNA technology. Now, that's about to change.

    Last month, Makerere University here in Uganda's capital opened its new Molecular Biology Lab, the first university center to provide a full complement of facilities for molecular biology, including polymerase chain reaction (PCR), cloning, and DNA sequencing. “This lab is a gold mine, because many excellent local students with an interest in wildlife are ready to jump at the opportunity of applying DNA technology,” says population geneticist Michael Bruford of the University of Cardiff in the United Kingdom, who has extensive experience in building research capacity in developing countries.

    Lab director Silvester Nyakaana, a population geneticist, says the facility will initially focus on studies of the phylogeny and population genetics of large African mammals. Researchers will use PCR to pull out and amplify highly variable genetic markers, which they will sequence to yield information on genetic divergence and the relationships between population groups. “Genetic studies carried out now are our last chance to gain a deep insight into how groups of large mammals organize and spread over time,” says Arctander, who studies the molecular evolution and population genetics of herbivores. “Africa is the only place where such groups still exist, and their distribution and genetic purity are being disturbed at a fast pace.”

    The Kampala lab is the product of a 6-year effort by Arctander and Panta Kasoma, director of Makerere's Institute for Environment and Natural Resources (MUIENR). Arctander's research has taken him to Africa many times over the past 8 years, and he visited MUIENR frequently, learning firsthand the frustrations of trying to conduct genetic studies without adequate facilities. He and Kasoma drew up a plan to establish a DNA facility at Makerere, and Arctander took the idea to a Danish government development program called ENRECA, Enhancement of Research Capacity in Developing Countries. ENRECA came through with a 9-year grant of $1.7 million, and the project was in business.

    Arctander began by training a group of young Ugandan scientists at his Copenhagen lab in the application of DNA technology, while a building on the Makerere campus commonly known as the cowshed was converted into a state-of-the-art genetics lab. This compact yellow bungalow now contains brightly lit labs and storerooms, decked out with sleek modern machinery and inviting Danish-designed lab furniture —a stark contrast to the dark and decades-old wood-paneled teaching labs found elsewhere on the campus. The ENRECA funding will also pay the first 3 years of salaries for the lab's initial staff of 10 scientists, students, and technicians.

    It has not all been plain sailing, however. Even after the official opening, workers were still installing an elaborate system of generators to protect against frequent power failures. “After a week of trying, we are still waiting for electrical adapters for the equipment, and we have come to realize that even a piece of plastic tubing cannot just be borrowed from a neighboring lab,” says Pia Friis, Arctander's research assistant. And the local bureaucracy can be trying. Imports tend to get stuck in the airport for weeks to months, as happened with the lab's new vehicle for fieldwork.

    The lab's newly trained researchers are now itching to get to work. Biologist Josephina Birungi plans to study speciation within groups of antelopes by comparing selected mitochondrial DNA sequences. Initial studies suggest that these data could challenge the current species definition within this animal group, because the traditional morphologically based boundaries between species, subspecies, and populations are blurred. Meanwhile, graduate student Vincent Muwanika is preparing to analyze the speciation of the warthog, which has never been studied genetically. “My studies should solve the long-standing argument as to whether warthogs are divided into subspecies,” says Muwanika, who also hopes to answer the question of whether a population of warthogs recently identified in Kenya belongs to a species believed to be extinct.

    The Makerere researchers also hope to put their skills to work in conservation efforts, such as helping guide the relocation of animals. Although conservation sometimes makes moving populations of animals desirable, geographically distant groups may have evolved and adapted to different environments in ways that make them less compatible with each other, even though they may appear morphologically similar. “The only way to avoid the detrimental effects of mixing different subspecies is to look at genetic elements,” says Nyakaana, who conducted an extensive characterization of the genetic divergence between elephant populations in Uganda, which he carried out while visiting Arctander's lab.

    Kasoma hopes to attract foreign investigators and collaborators to the institute to help create an intellectual center of excellence. Arctander adds that part of the motivation for the project was “making it possible for scientists worldwide to study African subjects in Africa, so that the data can be recorded and put to use here.” Visiting investigators may also provide an important source of support when the funding from ENRECA runs out in 3 years' time. So far, Nyakaana says researchers from surrounding countries as well as from Europe and the United States have expressed interest: “There is a deep-felt need for genetic information in relation to research originating in Africa, and we intend to fill the gap.”


    Bible's Bad Boys Weren't Such Philistines After All

    1. Michael Balter

    The discovery of an ancient shipwreck adds to archaeological evidence suggesting the Philistines were a worldly trading people

    ASHKELON, ISRAELThe discovery of two 8th century B.C. Phoenician ships loaded with wine amphoras off the southern part of Israel's coast, announced last week, was much more than a triumph for high-tech deep-sea archaeology (Science, 12 February, p. 929). It may also burnish the image of the Philistines, a people who occupied the territory of the Levant nearest to where the ships were found. Frequently portrayed as villains in the Bible—the giant Goliath slain by David was a Philistine, as were those who blinded Samson after he was betrayed by Delilah—the Philistines and how they came to the shores of the Middle East more than 3000 years ago are largely mysteries. Now the underwater discovery, together with years of painstaking excavations of Philistine cities on land, are beginning to reveal a picture of a cosmopolitan people who traded widely across the eastern Mediterranean.

    Where Goliath walked.

    Philistia was one of the Middle East's coastal trading powers some 3000 years ago.

    “The Philistines have been defined mostly by their enemies,” says Harvard University archaeologist Lawrence Stager, co-leader of the group that found the ships, which apparently sank en route to Egypt or Carthage. “We haven't really allowed them to speak for themselves.” Adds Seymour Gitin, director of the W. F. Albright Institute of Archaeological Research in Jerusalem, “There is a great body of evidence now to show that the Philistines were not philistines in the sense we define the word today.”

    At the height of their nearly 600-year-long civilization, Philistia consisted of five major cities: Ashkelon, Ekron, Gaza, Ashdod, and Gath, strung out along a swath of coastal plain that encompasses the sites of present-day Gaza and Tel Aviv. But although the Biblical accounts suggest frequent skirmishing with the Israelites inland to the east, there is little real evidence that the Philistines ever tried to conquer their Biblical foe. Indeed, according to non-Biblical texts, they seem to have been more concerned with steering clear of the two real superpowers in the area—Egypt in the west and, in the east, first Assyria and later Babylonia. And the archaeological evidence shows that Philistia put much of its energy into producing the kinds of goods that soon made it into a major trading partner of the surrounding empires.

    Since 1985, Stager's team has been excavating the extensive ruins of ancient Ashkelon, once Philistia's leading seaport. In the center of the Philistine city, the team found a large building containing wine presses, vats, and basins. These finds, combined with extensive remains of both Philistine and Phoenician wine jars uncovered at the site, lead Stager to conclude that Ashkelon was a major wine producing and exporting center. And at Ekron, an inland Philistine city northeast of Ashkelon, major excavations by Gitin and archaeologist Trude Dothan of Jerusalem's Hebrew University uncovered more than 100 olive oil production facilities, making Ekron the largest known producer of this valued commodity in the ancient world. That the Philistines did not keep the wine and olive oil all to themselves was suggested by the considerable archaeological evidence of contact with Phoenicia and Egypt at Ashkelon, including masses of Phoenician pottery and a cache of bronze bottles depicting Egyptian gods, the latter found in the winery complex itself.

    The latest discovery firms up those clues. Searching for additional evidence of trade between Philistia and other regions, Stager teamed up with Titanic discoverer Robert Ballard of the Institute for Exploration in Mystic, Connecticut, to scan the seabed adjacent to Philistia with sonar. Last year this survey suggested the presence of offshore shipwrecks, which the team videotaped last month with the aid of the robotic JASON submersible of the Woods Hole Oceanographic Institution in Massachusetts.

    Stager estimates that the Phoenician ships, which are stuck in the mud at a depth of 500 meters, date from between 750 and 700 B.C. based on the type of wine amphoras they were carrying. Located about 50 kilometers offshore, the ships are loaded with about 750 of the amphoras, possibly taken aboard at Tyre or some other Phoenician port, and may have stopped at Ashkelon before continuing onward. While the exact location of the ships is being kept secret, Stager told Science that they are roughly on a line between Ashkelon and the north Egyptian coast. Moreover, Stager says, their bows are both facing west. “This shows that trade from Phoenicia was coming across the Philistine coastal plain, on its way to Egypt,” says Gitin. “It's magnificent.” Another possibility, Stager says, is that the ships were headed for Phoenicia's colony of Carthage, near present-day Tunis. “It would have been a very nice cargo for the colonists,” he says.

    If Phoenician ships did regularly stop at Ashkelon, as suggested by the Phoenician pottery found in the Philistine city, it would support the growing consensus that the Philistines were a major trading nation in the region. “Ashkelon was a crossroads, a port on the sea route between Egypt and Syria,” says University of Vienna archaeologist Manfred Bietak, who for many years has been excavating the ancient Egyptian city of Avaris. “[The excavations at] Ashkelon will be an inexhaustible source of new surprises for many years to come.”

    Stager hopes they will also shed some light on where the Philistines came from—a question that is by no means settled. The Philistines are the best known of the so-called Sea Peoples, invaders who, beginning about 1200 B.C., swept eastward across the Mediterranean, sacking and pillaging as they went. Most scholars believe the invaders were originally Mycenaeans, whose culture spread from mainland Greece to Crete, Cyprus, and the Aegean coast of modern-day Turkey from about 1400 to 1100 B.C. According to this theory, the Mycenaeans displaced some of the local Levantine inhabitants, referred to as Canaanites in the Bible. Thus the earliest Philistine pottery is painted in a “monochrome” style typical of Mycenaean ceramic ware, although later Philistine pottery takes on a characteristic “bichrome” appearance, usually consisting of red and black lines and designs.

    At Ashkelon, Stager's excavation recently uncovered more evidence for this view: cylinders made of unbaked clay, which Stager thinks are loom weights designed to hold threads in place while they are being weaved, found on the floor of a 12th or 11th century B.C. building. “The Canaanites used pyramidal, pierced loom weights,” says Stager. “No one we know in the Near East used these cylinders, which could be used to spool thread, but they are found in textile contexts in coastal Cyprus, in Crete, and later in Greece … so it fits very nicely into the areas that we are considering for the [original] homeland.”

    To get a better fix on Philistine origins and the nature of the culture they replaced, the Ashkelon team is now excavating the Canaanite levels below the Philistine occupation strata. Last year, the archaeologists completed the excavation of a massive mudbrick gate and rampart apparently dating from the Canaanite era, and the excavations so far this season have penetrated a major Canaanite rock tomb from which the remains of at least 40 men, women, and children have been recovered. Somewhere between the well-defined Philistine and Canaanite levels, Stager says, “we think we will find the stage where, if they came by sea, we will be within a generation of the [Philistines] landing here.”

    Wherever they came from, the archaeological evidence is growing that the long-maligned Philistines were at the center of the international cross-currents of the time. But these heady days did not last long: In 604 B.C., the Babylonians, led by Nebuchadrezzar, put Ashkelon to the torch. Ekron suffered the same fate the following year and the other cities soon followed. In the winery at Ashkelon, Stager's team found smashed pottery, charred wood, and melted mudbrick, signs of the catastrophe that befell the once proud seaport.


    Lack of Icebergs Another Sign Of Global Warming?

    1. Bernice Wuethrich*
    1. Bernice Wuethrich is writer in Washington, D.C.

    Unlike the Titanic, the Queen Elizabeth II is the most conservative of cruise ships when it comes to icebergs, if need be, sailing far south to avoid the bergs that normally pock the North Atlantic. But this spring the luxury liner crossed the notorious Iceberg Alley without a second thought. For reasons no one understands—although global warming is a top suspect—the Grand Banks shipping lanes, which are located southeast of Newfoundland, were an ice-free zone.

    For the first time in 85 years, the International Ice Patrol (IIP) issued not a single bulletin reporting lurking bergs. “The lack of ice is remarkable,” says IIP Commander Steve Sielbeck: The Coast Guard's IIP has been tracking icebergs that wander south of 48°N latitude ever since the Titanic sank in 1912. And by now the trackers thought they knew what to expect. In an average year, some 500 bergs drift down the Labrador Coast from western Greenland, where most are spawned.

    But the number varies widely, depending on a quasi-decadal cycle of high and low atmospheric pressures known as the North Atlantic Oscillation (NAO). High NAO years typically mean strong northwesterly winds that bring cold Arctic air to the Labrador Sea—and push convoys of icebergs toward the shipping lanes. Low NAO years usually coincide with low iceberg frequencies. This year, however, the NAO was high, but the winds inexplicably blew in from the northeast. “Something is out of sync,” says Ken Drinkwater, an oceanographer at the Bedford Institute of Oceanography in Dartmouth, Canada.

    The northeasterly winds stranded many icebergs against the Labrador coast, and because of unusually warm water and air, fewer than usual had drifted south in the first place. John M. Wallace, a meteorologist at the University of Washington, Seattle, notes that since the 1980s, winter temperatures have risen at least 0.5°C poleward of 45° north, a line that runs through the Grand Banks. Until now, the chilling effect of westerly winds had masked this warming, which Wallace attributes to the general warming of the globe. This year, he says, “The decline of westerlies and global warming are working together.”

    The warmth melted sea ice, the frozen sea water that buffers icebergs from wave erosion and warmer water. This spring sea ice in the region was about as scarce as it has ever been, says Simon Prinsenberg, a research scientist at Bedford. Furthermore, water in the Grand Banks itself was 2°C above normal—warm enough to instantly melt any remnant ice that kissed its border.

    Despite all the heat, nobody is totally willing to discount the possibility that some kind of natural climate fluctuation is at play, albeit in a more extreme form than usual. Will the icebergs return next year? “Without a doubt,” Sielbeck says. “I just can't tell you how far south they'll get.”

Log in to view full text

Via your Institution

Log in through your institution

Log in through your institution

Navigate This Article