# News this Week

Science  18 Jul 1997:
Vol. 277, Issue 5324, pp. 312
1. EXOTIC DISEASES

# Is an Old Virus Up to New Tricks?

1. Jon Cohen

A RECENT UPSURGE OF MONKEYPOX CASES IN THE DEMOCRATIC REPUBLIC OF CONGO, IN WHICH THE VIRUS SEEMS TO HAVE JUMPED FROM PERSON TO PERSON MORE READILY THAN BEFORE, HAS RESEARCHERS ON FULL ALERT

Exotic viruses like hanta, Ebola, and Marburg have become media stars in the past few years, the focus of blow-by-blow news accounts, best-selling books, and doomsday Hollywood thrillers that chart attempts by scientists to thwart the onslaught of these “emerging” pathogens. Yet little has been heard outside the scientific literature and Internet chat groups about an exotic infection that is alarming some public health experts: the largest outbreak ever seen in humans of a well-known virus called monkeypox. A first cousin of the once-dreaded smallpox, monkeypox causes nearly identical symptoms. “I hate to be accused of pushing the alarmist button, but for practical purposes, smallpox is back,” says virologist Peter Jahrling of the U.S. Army Medical Research Institute of Infectious Diseases in Fort Detrick, Maryland.

Although researchers have long known that monkeypox routinely jumps from animals to humans, studies have shown that infections quickly die out, because the virus does not easily pass from one person to another. Yet between February 1996 and February 1997, the Democratic Republic of Congo (formerly Zaire) had at least 92 cases of the disease and three deaths in a remote region where only 37 documented cases were reported from 1981 to 1986. These findings have led to intensive discussions about what might be causing the outbreak—in particular, whether the virus is traveling from human to human more readily than before. The outbreak also has stoked the debate about the pros and cons of retaining samples of the smallpox virus—which has no animal hosts and only exists in two lab freezers—for comparative research with related viruses like monkeypox (see sidebar).

Medical epidemiologist Ali Khan of the U.S. Centers for Disease Control and Prevention (CDC) in Atlanta became seriously worried about the outbreak spreading after visiting the area in February. Khan headed an international team, organized by the World Health Organization (WHO) and the Zairian Ministry of Health, that visited the 12 villages at the center of the outbreak (see map) for 5 days; the trip was cut short because of civil unrest. As Khan and colleagues detailed in the CDC's Morbidity and Mortality Weekly Report (MMWR) on 11 April, they believe 73% of the 89 people they studied were infected by other people. That would be a large jump from the 30% “secondary contact” rate reported by Czechoslovakian epidemiologist Zdenek Jezek, who did the 1981 to 1986 studies in Zaire and co-authored the book on the disease, Human Monkeypox. The recent investigation also found one patient who appeared to have been the source of eight other infections, which is twice the highest chain of transmission observed previously. “This is not just an outbreak of some rare, exotic disease in the middle of nowhere,” says Khan. “I'm personally concerned about what would happen if this disease showed up in a major city.”

Joel Breman, who formerly headed a monkeypox effort as part of the WHO's smallpox eradication program, says he is “alarmed” by the number of cases in such a small area, the secondary attack rates, and the data on chain of transmission. But Breman, who is now with the Fogarty International Center at the U.S. National Institutes of Health, isn't convinced that human-to-human transmission is increasing. “They really were not there long enough to do a detailed ascertainment of what's going on,” says Breman. “I would not want to jump to any conclusions that this is new and radically different from what was going on before.” Jezek, a retired consultant with WHO, agrees: “I've seen human-to-human transmission, and it's very difficult to distinguish whether it's an introduction from an animal source.”

## Success breeds vulnerability

One explanation for the outbreak, floated by Jezek and others, raises a bitter irony: Smallpox vaccinations—which also confer immunity to monkeypox—were so successful that they were stopped nearly 2 decades ago, creating an ever-growing pool of people who are susceptible to the animal virus. Indeed, only 18% of the people studied in the recent outbreak had the characteristic smallpox vaccination scar on their upper left arms. Moreover, even those who were vaccinated may no longer be protected, because, says Jezek, immunity wanes after 5 to 10 years. Immunity is further compromised by this country's high rates of infection with HIV, which cripples immune systems.

The civil war may have contributed another factor: Faced with a growing threat of starvation because of the unrest, villagers may have increased their hunting of animals that carry monkeypox, which include monkeys, squirrels, and rats. “If the population has no sources for food, they start to hunt what is very close,” says Jezek. “On the palm trees surrounding the villages, there are plenty of squirrels.” Khan and colleagues attempted to investigate this possibility during their recent trip, collecting blood samples from 64 squirrels (one of which had a suspect lesion), rats, and other animals. These samples still are being analyzed.

A more frightening hypothesis is that monkeypox really has changed, becoming more virulent or more transmissible. But, as reported in the MMWR article, researchers have genetically analyzed part of one isolate from a person infected in 1996, and there's no evidence that it differs from strains collected in Zaire between 1970 and 1979. In a 17 May letter in The Lancet, Khan and several of his international colleagues emphasize this point, stressing that “notions of monkeypox virus mutating into [smallpox] virus are unfounded.” Researchers are, however, concerned that a monkeypox variant could have found its own way to become more adapted to the human host—which would be an ominous prospect. “This could be worse than smallpox if it adapts to humans,” acknowledges virologist Bernard Moss of the National Institute of Allergy and Infectious Diseases (NIAID). “Then we'd have smallpox with an animal reservoir. How's that for a scenario?” Fortunately, there is no evidence so far to support such a scenario.

The Army's Jahrling thinks “the probability is low” that this outbreak will spread further. “But it's not zero,” he adds. Because of that possibility, Jahrling convinced the military to give U.S. troops the smallpox vaccine if they should be ordered into this region. But researchers are reluctant to recommend a new vaccination program for the local population, because the vaccine can cause disease and death in people whose immune systems have been decimated by HIV. “The number and the frequency of side effects from the vaccine would be much higher than the cases of monkeypox,” predicts Jezek.

That quandary led the CDC's Joseph J. Esposito, who co-authored the Lancet paper and since 1973 has studied the Orthopoxvirus family—to which monkeypox belongs—to the idea of giving these vulnerable populations a weakened version of the smallpox vaccine, known as MVA. But it's unclear whether MVA, which decades ago was tested extensively in Germany against smallpox, would protect against monkeypox. He would now like to test that hypothesis in monkeys. Esposito says there is also the chance a new drug might help. In collaboration with U.S. Army and NIAID researchers, he has been studying drugs that are already in human trials for other diseases and, in test tube experiments, appear to have antiviral activity against monkeypox, cowpox, and even smallpox (which he determined by tests with three smallpox strains in storage at CDC).

For now, health authorities are trying to keep an eye on the outbreak. The United Nations Integrated Regional Information Network put out a bulletin on 20 March warning that the retreating Zairian soldiers—who were responsible for Khan and colleagues fleeing the country—were moving through the very region where the outbreak occurred. The bulletin also noted that aid workers in the region were worried that Rwandan refugees in poor health were heading toward the region and “could contribute to an escalation in the number of cases and its spread.”

To get a closer look, WHO hopes to organize another mission in a month if the political situation calms down enough. Khan is anxious to return. “It's important to get back out there and ask ‘What's the magnitude of disease?’ and ‘How does transmission occur?’ “says Khan. “We're all waiting for better data.”

2. EXOTIC DISEASES

# Smallpox: Clues From a Killer

1. Jon Cohen

Nearly 4 years ago, Science ran side-by-side Policy Forums, written by groups of prominent researchers, arguing for and against destroying the world's only remaining stocks of smallpox virus (Science, 19 November 1993, pp. 1223 and 1225). A main argument made by the proponents of keeping the two stocks—which reside only at the U.S. Centers for Disease Control and Prevention (CDC) in Atlanta and at the Research Institute for Viral Preparations in Moscow—was that studies of smallpox might help combat another member of the Orthopoxvirus family, monkeypox. Now the monkeypox outbreak in the Democratic Republic of Congo (see main text)—which has yielded more cases than ever before, and more human-to-human transmissions—has strengthened the convictions of those researchers. “The fact that monkeypox outbreaks are still occurring, and are not going to be eliminated, is a strong argument for retaining [the stocks],” says Duke University's Wolfgang Joklik, first author of that editorial.

Sergei Shchelkunov of the State Research Center of Virology and Biotechnology in Koltsovo, Russia, is now sequencing the complete genome of monkeypox. Once that is completed, he argues that researchers will need live smallpox virus (which already has been sequenced) to test their new ideas about how the poxviruses cause disease. “I believe we will carry out some laboratory experiments in the future with smallpox virus,” contends Shchelkunov, who is collaborating with a co-author of the save-smallpox editorial, Bernard Moss of the National Institute of Allergy and Infectious Diseases. “Without this virus, it will be impossible to get that extremely important information.” For example, several researchers say the only way to tease out why monkeypox is less infectious to humans than smallpox—to determine if the viruses use different receptors to enter cells, for example—is with live viruses.

Yet virologist Joseph J. Esposito of the CDC, another collaborator on the monkeypox sequencing project, doesn't think the remaining smallpox is needed to combat monkeypox. “I don't think monkeypox is going to fill a niche that smallpox left behind,” says Esposito. “We're pretty far advanced, and we can bring these viruses under control.”

Zdenek Jezek, co-author of the main book on monkeypox and a former World Health Organization consultant, sees eradicating the smallpox virus simply as a necessary last step that must be taken, and this outbreak changes nothing. “Keeping smallpox is something like a stamp collection,” says Jezek. “People are afraid it will disappear from their collections.” Plans call for destroying the remaining isolates on 30 June 1999, although a World Health Assembly meeting the month before could stay the execution.

3. BIOCHEMISTRY

# Yeast Protein Acting Alone Triggers Prion-Like Process

1. Gretchen Vogel

The notion that a misshapen protein can transmit disease—a feat long thought to be the exclusive preserve of living pathogens—has moved from the heretical to the mainstream in the past decade. A growing body of circumstantial evidence has convinced many researchers that rogue proteins, called prions, are the villains behind diseases like Creutzfeldt-Jakob disease (CJD) in humans and “mad cow disease” in cattle. Yet nobody has produced a smoking gun: clear proof that a prion protein can wreak havoc without an accomplice.

Researchers studying a similar phenomenon in yeast are getting close to a conviction, however. On page 381, molecular geneticist Michael Ter-Avanesyan and his colleagues at the Institute of Experimental Cardiology in Moscow report that in the test tube, a purified protein can cause changes in other protein molecules, converting them into an insoluble lump of proteins like those seen in affected yeast. The scientists think the twisted molecule acts as a template, inducing others to change shape—the same process by which prions are theorized to cause disease in mammals.

The evidence is still not watertight, but the work is a huge leap forward, says Reed Wickner, a geneticist at the National Institute of Diabetes and Digestive and Kidney Diseases, who first proposed that a prion-like process might explain why two types of yeast, called [PSI+] and [URE3], are able to pass on a new trait to their offspring without changing their DNA. Biochemist Byron Caughey, who studies mammalian prions, agrees. “They have done a lot of things that we would hope to do [in mammals],” he says. And, although it's a big leap from the test tube to living cells—and an even bigger one from yeast to mammals—supporters of the prion hypothesis say the growing consensus among yeast researchers strengthens their case.

Ter-Avanesyan's work focuses on a protein called Sup35 that normally helps the yeast proofread during the DNA translation process, which converts the DNA code into proteins. In [PSI+] strains it is inactive, allowing the yeast to make proteins from certain stretches of DNA that would otherwise be ignored as unintelligible. A year ago, yeast biologist Susan Lindquist and her colleagues at the University of Chicago showed that in living cells, the Sup35 protein is evenly distributed in normal [psi] cells but is clumped together in [PSI+] cells. At about the same time, Ter-Avanesyan and his colleagues showed that the clumping inactivates the protein, and that it takes place when Sup35 proteins stick together at one end. That still left a question, however: What caused the clumping?

Ter-Avanesyan and his colleagues set out to show that the [PSI+]'s Sup35 protein alone—not a combination of proteins or a more complicated series of events—was the culprit. They combined cell extracts from normal [psi] yeast, in which Sup35 is soluble, with cell extracts from [PSI+], in which Sup35 is insoluble, and then spun the mixture in a centrifuge to separate the soluble and insoluble proteins. In as little as 20 minutes, much of the soluble protein had become insoluble, indicating that the protein had clumped together. After 2 hours, no Sup35 was left in the cell solution; it was all in the insoluble pellet.

To eliminate the possibility that, in [PSI+] cell extracts, something other than Sup35 triggered the clumping, Ter-Avanesyan and his group purified just the sticky end fragment of the protein and added it to a solution of normal Sup35. The protein duly formed a clump. The researchers then used a piece of the newly formed pellet to convert a fresh solution of normal Sup35. By the third pass, the signature of the original protein fragment had disappeared from protein-detecting blots, suggesting that the newly converted proteins can indeed pass along their new conformation.

These results fit in well with two other recent studies of Sup35. In the 30 May issue of Cell, Lindquist and her colleagues showed that Sup35 produced in genetically engineered Escherichia coli can spontaneously form long fibrils in the test tube—similar, at first glance, to the fibrils present in the brains of Creutzfeldt-Jakob disease patients and bovine spongiform encephalopathy (BSE)-infected cows. They also showed that adding the fibrils to freshly prepared solutions of the protein caused new fibrils to form much more quickly than in “unseeded” solutions. And in the 24 June Proceedings of the National Academy of Sciences, biophysicist Kurt Wüthrich and molecular biologist Chih-Yen King of the Institute for Molecular Biology and Biophysics in Zurich and their colleagues showed that Sup35's sticky end alone also will form long fibrils in the test tube. In addition, they report two lines of evidence that a shape change is responsible for the fibril formation. While the freshly prepared protein solution shows little evidence of specific shape under polarized light, the fibrils look like they might contain β sheets, the same conformation that is suspected in mammalian proteins. In addition, when the researchers stain their fibrils with Congo Red—a stain used to identify characteristic plaques in the brains of CJD patients—they found that, like CJD plaques, the stained aggregates appear yellow-green under polarized light, a characteristic of an ordered pattern in the protein aggregate.

So far, researchers studying mammalian prion proteins haven't been able to conduct the kinds of serial experiments Ter-Avanesyan carried out. One reason is that so much of the abnormal mammalian protein appears to be required to trigger changes in the normal version that it is impossible to separate out just the newly converted protein to see if it can, in turn, convert a fresh batch of normal proteins.

But at least one longtime critic of the idea that infectious proteins can cause disease praises the new work in yeast. Yale neuropathologist Laura Manuelidis, who thinks CJD and BSE are caused by an as yet undetected virus, reported in the 4 July issue of Science (p. 94) that she and her colleagues were able to induce a BSE-like disease without any evidence of abnormal PrP protein. She says the new work is interesting for its insights into how proteins may be part of the disease process, but cautions that showing that proteins can induce others to change shape in the test tube is very different from showing that they can cause infectious disease.

Even in yeast, researchers acknowledge that the protein-only case is not yet closed. Doubts will remain until someone is able to insert purified protein into a living cell and show that it causes the [PSI+] trait. “That's the final nail, but that little hole is still there, and it needs to be filled in,” says Lindquist, adding: “There's the same Holy Grail as for the mammalian prion story.”

4. ECOLOGY

# Resurgent Forests Can Be Greenhouse Gas Sponges

1. Anne Simon Moffat

When people talk about combating global warming, they usually mean cutting back on the burning of fossil fuels or on slash-and-burn agricultural practices in the tropics. But ecologists and other researchers are realizing that there's another way to control the carbon dioxide build up that may be warming the planet: plant trees.

The strategy grows out of recent evidence suggesting that forests store much more carbon than had been thought. Previous estimates indicated that they take up about as much carbon dioxide while photosynthesizing as they give off when respiring—resulting in little net carbon flow into or out of forests. But new results, some from reanalyses of old data and others from field studies, indicate that “forests, and the carbon they sequester, have been undervalued,” says Harvard University environmental economist Theodore Panayotou.

Earlier studies had neglected to include the huge amounts of carbon stored in peat and other organic matter in soils—now estimated to account for about two-thirds of the total sequestered, primarily in high-latitude forests. In addition, contrary to popular belief, many forests are expanding, which also helps draw carbon dioxide out of the atmosphere and lock it away in organic matter.

This new picture of forest dynamics may help solve a long-standing puzzle. When researchers estimate annual carbon dioxide releases and compare those figures to actual carbon dioxide concentrations in the atmosphere and known sinks, such as the oceans, they typically come up short by about 1 billion to 2 billion metric tons. In other words, roughly 20% of the total released each year is apparently missing. But the forest studies suggest where at least part of this carbon dioxide could be going. “Terrestrial systems, including forests and their soils and agriculture, can account for some of the missing carbon,” says ecologist Steven Hamburg of Brown University.

On the practical side, the findings have led to the idea that better forest management techniques, including, for example, stemming tropical deforestation and replanting logged areas, could lead to even greater removal of carbon dioxide from the atmosphere. Indeed, some electrical utilities have already begun to put such strategies into practice as a way of compensating for their emissions. “In the last year or two, we've gone from modeling to real experiments,” says ecologist Robert Dixon, who directs the U.S. Country Studies Program, a federal interagency activity that inventories greenhouse gas emissions and identifies ways to reduce their impact.

No one believes that such practices alone can stem escalating carbon dioxide emissions. But a special working group within the Intergovernmental Panel on Climate Change, which is putting together ideas to manage global climate change, suggests that improved use of forests worldwide could sock away enough carbon in soils, trees, and other vegetation between 1995 and 2050 to offset 12% to 15% of fossil fuel emissions during that period. “Generally speaking, there is agreement that forestry is one piece of a set of mitigation strategies that should be pursued,” says analyst Mark Trexler, president of a climate-change mitigation consulting firm in Portland, Oregon.

Data suggesting the strategy's potential began building about 5 years ago. In a study reported then, forest researcher Pekka Kauppi and his colleagues at the Finnish Forest Research Institute in Helsinki assessed the growth in European forests from 1970 onward (Science, 3 April 1992, p. 70). They started with data in the various national forest inventories—information that had long been neglected, Kauppi says, because it was published in the “gray literature,” the agency reports of different governments. But, he adds, “the quality of these data is often very high.”

Analysis of those data showed that each year during the 1970s and 1980s, European forests accumulated 70 million to 105 million tons of carbon, with an additional 15 million tons tied up annually in wood that was harvested and used in construction. Based on these figures, Kauppi and his colleagues calculated that the carbon picked up by European forests could account for 8% to 10% of the missing carbon dioxide worldwide.

Some of that carbon accumulation was due to growth of existing forest vegetation, but reversion of agricultural lands to forests played a bigger role—a finding since bolstered by more recent studies, such as one completed just last year by forest scientist Richard Birdsey of the U.S. Department of Agriculture Forest Service Laboratory in Radnor, Pennsylvania, and his colleagues.

They found that over the last 100 years, between 9 million and 11 million hectares of agricultural lands in the eastern and southern parts of the United States have reverted to forests. Indeed, Birdsey calculates that the increase in biomass and organic matter on U.S. forest lands over the last 40 years has stored enough carbon to offset 25% of U.S. greenhouse gas emissions for that period. In addition, Birdsey's projections show there is enough land available for further forestry to increase carbon sequestration by perhaps up to 40%, at least until 2040.

Taking a more global view, Dixon, forest ecologist Sandra Brown of the University of Illinois in Urbana, and others completed the first inventory of the carbon stored in all forests about 3 years ago. They calculate that worldwide, temperate forests, because they are young and still growing, sequester about 0.7 billion metric tons of carbon annually. That figure is more than offset by the 1.6 billion metric tons released each year by deforestation in the tropics. But it suggests that good management could turn forests into major carbon “sinks.”

Much of the carbon stored by these temperate forests is not in trees, shrubs, or other aboveground vegetation. Dixon and Brown's survey showed that only about one-third is in vegetation; the other two-thirds is in soils, much of it in peat, especially at high latitudes. These results help explain why previous estimates of the carbon sequestered by forests were low: They did not take into account the large amount of carbon in peat.

But the capacity of these high- or mid-latitude forests to store carbon may pale compared to that of tropical rain forests, according to an analysis now being carried out in Panama, Malaysia, and nine other tropical locales by a team from Harvard University and the Smithsonian's Center for Tropical Forest Science. These researchers, including Harvard's Panayotou and the Smithsonian's Elizabeth Losos, are monitoring the growth of every tree in a series of 50-hectare plots and recording their responses to catastrophes, both natural and those caused by human activities, such as logging. These and other studies show that tropical forests can sequester as much as 200 metric tons of carbon per hectare, says forest economist Marco Boscolo of Harvard. This is about one and one-half to two and one-half times the storage capacity of forests of mid- and high latitudes.

The problem will be in realizing the vast sequestration potential of the tropics, given the ongoing deforestation there. Still, there are signs that deforestation can be curbed, if governments make the necessary commitment. “Although reducing deforestation in the tropics may appear to be difficult, it is happening in countries such as Brazil, India, and Thailand,” says Brown. Strong forest legislation, a large reforestation program, and community awareness are responsible for the turnaround, she says.

In some cases, electrical utilities are helping out, partly because it may help them win the permits they need to expand production. Because carbon dioxide can travel around the world, U.S. utilities can be recognized for reforestation done abroad. For example, a group of utilities including Wisconsin Electric Power and the PacifiCorp is spending $3.5 million to protect about 2500 hectares of forest in Belize. And in Sabah, Malaysia, that state's largest logger, the Innoprise Corp., is cooperating with the Forest Absorbing Carbon Dioxide Emissions Foundation, a group started by the Dutch Electricity Generating Board, to rehabilitate 25,000 hectares of degraded, logged forests with a mix of long-lived local trees, known as dipterocarps, and forest fruit trees. The goal is to increase the carbon sequestered in those areas by at least 200 metric tons of carbon per hectare. Similar projects are under way at home. For example, the Klamath Cogeneration Project, which beat out two competitors to build a new, gas-fired power plant in Klamath Falls, Oregon, plans to reforest 1000 hectares of Oregon grassland, pasture, and scrub—at a cost of$1.5 million—to offset some of the plant's carbon dioxide emissions.

Even as these efforts are getting started, researchers are beginning to test novel ideas about how forests can be managed to soak up even more carbon. A computer-modeling effort by Gregg Marland and his colleagues in the Environmental Sciences Division of the Oak Ridge National Laboratory in Tennessee indicates, for example, that actively managed, fast-growing forests that are harvested and replanted can tie up even more carbon than mature, protected forests that are left alone. That's because harvesting wood for use in long-lived products, such as home construction, takes the carbon out of circulation, in effect creating a new carbon sink. Moreover, wood fuels can substitute for fossil fuels, such as coal and oil, that throw more carbon dioxide into the atmosphere.

The Marland team's results are buttressed by the results of a study in which J. Piers Maclaren, Steve Wakelin, and Lisa Morenga of the New Zealand Forest Research Institute Ltd. analyzed that nation's plantations of Pinus radiata. They found that a radiata pine plantation that is replanted each time it is harvested puts away 112 metric tons of carbon per hectare by the time the trees mature. “This,” says Maclaren, “can be seen as a one-time, permanent movement of carbon from the air to the land surface.”

Maclaren calculates that from 2008 to 2026, New Zealand could more than offset its carbon emissions from fossil fuel combustion by intensive forest planting, at a rate of about 100,000 hectares per year. However, carbon sequestration would decrease rapidly after 2045, when plantable land was used up.

Indeed, forestry can go only so far in offsetting global warming, because land for new forests will eventually run short. Kauppi suggests that well-managed forests may serve as a carbon sink only through the end of the 21st century, while Brown sees forests filling a role as major carbon sinks for at least several centuries. Whatever the limits, she says, “it's an opportunity that must be pursued.”

5. ECOLOGY

# Hybrids Consummate Species Invasion

BOULDER, COLORADO— When biologists think of the comings and goings of species, they often think of war—of new species invading and pushing out the old. In midwestern lakes, however, love seems to be the driving force. In lake after lake, an invading crayfish species is pushing local crayfish to extinction. But biologists at the University of Notre Dame in Indiana are finding that the local crayfish are having their own effect on the invader, as the two species produce a new population of vigorous hybrids.

The finding is a surprise, researchers say, because ecologists often expect animal hybrids to be sterile, unable to play more than a bit part in species invasions. But at the annual evolution and natural history meetings here,* William Perry, a graduate student in the labs of ecologist David Lodge and biologist Jeff Feder, described molecular studies showing that hybrids of Kentucky native Orconectes rusticus, or the rusty crayfish, and a native crayfish, O. propinquus (the blue crayfish), are indeed fertile. Other work by Perry, Lodge, and Feder suggests that these hybrids are outcompeting both natives and invaders. The rusty crayfish, it appears, is taking over by assimilation.

That's useful information for conservation, notes Christopher Taylor, a crayfish systematist at the Illinois Natural History Survey, because rusty crayfish and similar intruders are slowly pushing natives to extinction: Of the 340 species of crayfish found in North America, about 30 may soon be completely eliminated by invaders, which usually arrive in new lakes as bait brought by anglers. “If we know how they are doing it,” says Taylor, “maybe we can think of a way to slow them down.”

Researchers noticed 2 decades ago that some Wisconsin crustaceans are intermediate in color and in the size of various body parts between the larger rusty crayfish—which first appeared in northern lakes in the 1960s—and the blue crayfish. That suggested to Lodge and his colleagues that invading and local species sometimes interbreed. But rusty crayfish themselves vary greatly in form, making it difficult to identify hybrids reliably. So the extent of hybridization remained unclear, and in any case, hybrids were assumed to be less important than other species-replacement mechanisms.

But when Perry collected specimens from some Wisconsin lakes and analyzed enzymes that serve as distinctive species markers, he found that extensive hybridization is under way between rusty and blue crayfish. Further comparisons revealed that backcrosses between hybrids and rusty crayfish were nearly as common as first-generation hybrids, indicating that hybrids are fertile and that they tend to mate with rusty crayfish rather than with each other. Together, the first-generation hybrids and backcrosses accounted for 30% of the crayfish in one lake.

From laboratory observations, Lodge and his colleagues had thought that most of the interspecies matches would be between the large, aggressive rusty males and the blue females. But when Perry examined the hybrids' mitochondrial DNA—which is inherited only from the mother—he found, to his surprise, that 89% were offspring of the opposite match, between rusty females and native blue males. “We were looking for love in all the wrong places,” he quips.

The apparent prowess of the hybrids may be speeding the invasion. When Perry put rusty and blue crayfish in tanks with similarly sized hybrids, the hybrids beat both species in competition for food—such as insects and aquatic plants—and for shelter under rock piles. “They are actually more competitive than the invader,” says Perry. “They're pretty nasty.” The crayfish invasion may start with love, but it ends up in war after all.

• * Joint meetings of the Society for the Study of Evolution, the American Society of Naturalists, and the Society of Systematic Biologists, Boulder, Colorado, 14–18 June.

6. MEETING BRIEFS

# How Male Animals Gain an Edge in the Mating Game

1. Elizabeth Pennisi

College Park, MarylandIn the midst of the East Coast's first summer heat wave, some 600 biologists gathered on 21 to 26 June here at the University of Maryland for the annual meeting of the Animal Behavior Society. Among the many talks and symposia were provocative discussions of the lengths male animals have gone—evolutionarily speaking—to find a mate.

## Croaking With All Ears

“Jug-o'-Rum,” bellows the American bullfrog on many summer evenings to court his mate. The call has long been thought to radiate from the large vocal sac that bulges from under the frog's chin. But Alejandro Purgue, who studies animal acoustics at the University of California, Los Angeles (UCLA), has traced it to a different part of the bullfrog's anatomy: its ears.

“They are acting as loudspeakers,” amplifying the sound of the frog's vocal cords, says UCLA neuroethologist Peter Narins. “I would never have guessed that in a million years,” says Philip Stoddard, a neuroethologist at Florida International University in Miami. And as Narins reported at the meeting, the bullfrog isn't the only frog that speaks through its ears.

Purgue discovered this unexpected use for bullfrog eardrums while trying to pin down how the frogs use their vocal sacs. As he will describe in an upcoming issue of the Journal of Comparative Physiology, he designed a sound-generating device that could be placed inside a frog's mouth to provide a reproducible sound for each test. He then turned on the sound and measured how various body parts amplified it. The vocal sac did pick up the vibrations, but a much stronger response came from the eardrums.

These measurements imply that the frog not only hears, but also calls, through its ears. The ears account for 70% to 80% of the sound output, he estimates, a finding Stoddard calls “absolutely stunning.” The sac, for its part, serves primarily to store the air used by the vocal cords, says Purgue.

His finding helped Narins begin to make sense of another frog, Petropedetes parkeri, which he had collected in Cameroon to study its odd-shaped ear. Based on a textbook picture, Narins had thought the frog had an enlarged middle ear bone. Not until he and his colleagues caught one did he realize that the protuberance, called a papilla, wasn't a bone at all. “It feels like a sponge,” says Narins.

The papilla develops only on males and only during the mating season. Based on preliminary acoustic analyses, Narins and his colleagues now think it tunes the eardrum to resonate at the dominant frequency in the frog's call, amplifying it. The dominant frequency in this frog's call is slightly lower than the natural resonating frequency of the eardrum itself. But “when the papilla is there, [the eardrum] matches the call more closely,” says Purgue.

“It looks like it's a unique adaptation,” says Andrea Simmons, a neuroscientist at Brown University. She wonders, however, whether it might not be used for a more traditional purpose: tuning the frog's hearing to specific frequencies. “It might be an aid to hearing,” Narins agrees. “But we haven't tested that yet.”

## Bowers as Barriers

Whereas peacocks court by flashing their elaborate tails, male bowerbirds win females by dancing on and around their bowers—platforms or towers made of sticks. Biologists who study these natives of Australia and New Guinea have long thought that the bowers help to wow the females by demonstrating the males' architectural prowess. But evolutionary biologist Gerald Borgia of the University of Maryland, College Park, reported that they serve a different purpose. His data suggest that bowers can provide a kind of screen, allowing the dancing and preening to be more vigorous without frightening off the female.

By correlating features of the bowers with the intensity of the males' displays, says Richard Prum, a systematist at the University of Kansas, Lawrence, Borgia found that “certain bowers have evolved in certain ways [to] allow males to display in a more energetic fashion.” Thus, the male spotted bowerbird (Chlamydera maculata) makes an unusually wide bower with see-through straw walls and approaches the female from behind them. “He's using the wall as a barrier,” says Borgia. In contrast, in two species that build relatively small platforms, “the display is much toned down.”

## All for Sperm's Sake

It's hard to miss a rooster crowing for his hens or a stag proudly displaying his antlers. But males also compete in more subtle ways to get their genes into the next generation, as their sperm vie to fertilize the female's eggs. Extensive studies of a common insect pest, the Indian meal moth, are now giving scientists a glimpse of the lengths males may go to give their sperm an edge. “The more I look, the more I find sophistication,” says Matthew Gage of the University of Liverpool in the United Kingdom.

At the meeting, Gage described how moths can size up females and beef up their sperm counts accordingly. They can also change the size of adult body parts to optimize their sperm's chances of success. Scientists have studied sperm competition since 1970, but Gage's studies show how thoroughly sperm competition can shape the rest of the male's life, says Gerald Wilkinson, an evolutionary biologist at the University of Maryland, College Park. “There are life history consequences to male reproductive investment, just as there are in females,” he says. “That hasn't been well appreciated.”

Gage raised larvae in jars with one, five, 10, or 20 individuals per jar, adjusting the amounts of food accordingly. When the larvae emerged as adults, they were all about the same overall size but were proportioned quite differently, depending on the population of the jars, he reported. Those raised alone or with only four companions had small testes and produced small amounts of ejaculate, presumably because they somehow sensed that their sperm would face little competition from other males mating with the same female. But they had relatively large heads and thoraxes—which contain the muscles and sensory systems needed to find mates. The moths from uncrowded jars also developed faster and lived longer, Gage notes, presumably because their lower investment in sperm meant they could afford to stock up more fat reserves to sustain the nonfeeding adult stage.

In contrast, adult males from the more crowded jars—where they might have easier access to females but lots of sperm competition—had smaller heads and thorax muscles. Instead, they had larger abdomens, presumably to enhance their sperm-generating capacity, and they provided females with larger packets of sperm. But this investment apparently came at a cost: These insects tended to die sooner than their counterparts in uncrowded jars.

Gage found that both uncrowded and crowded males also have another stratagem to give their sperm a boost. “They are capable of manipulating the material they are putting into the female,” Wilkinson notes. They make two kinds of sperm, one that can fertilize the egg and one that cannot because it lacks DNA. Gage suspects that the nonfertilizing sperm act as “cheap filler,” giving the female the sense that she has plenty of sperm and doesn't need to mate again. He found that males add more filler, plumping up their sperm contribution, when the receiving female is young and likely to have more mating opportunities.

In addition, all males insert a larger packet of sperm when the female is bigger, possibly to get her to wait longer before she mates again, or because her size indicates she has more eggs to fertilize, Gage suggests. “Sperm competition is an incredibly potent and far-reaching force, so it's not surprising that males have evolved cryptic, but sophisticated, systems to be successful,” he concludes.

7. PLANETARY SCIENCE

# Comet Origin of Oceans All Wet?

1. Donald Goldsmith

Blois, France Comets are made largely of ice, and over geologic history, vast numbers of them have hit Earth. Even now, according to a controversial proposal by University of Iowa, Iowa City, physicist Lou Frank, miniature comets might be quietly showering our planet (Science, 30 May, p. 1333). Many planetary scientists have surmised that most of the water on Earth's surface could have originated in comets. But a close look at the water in two recent comets challenges that conclusion.

At a planetary science meeting held here last month, Tobias Owen of the University of Hawaii's Institute for Astronomy reported that he, Roland Meier, and their colleagues had measured the ratio of ordinary water molecules to molecules containing deuterium, a heavy isotope of hydrogen, in this spring's spectacular comet, Hale-Bopp. The group picked up radio emissions from the deuterated water with the James Clerk Maxwell Telescope on Mauna Kea. The intensity of the emissions, the Owen-Meier team found, indicated that Hale-Bopp contains about three deuterium atoms for every 10,000 atoms of ordinary hydrogen. That's about twice the ratio in seawater. But it agrees with measurements of last year's comet Hyakutake, made by Daniel Gautier of Meudon Observatory and his colleagues and also announced at the meeting, and with the decade-old observations from the flyby of comet Halley.

“These results show that you can't make the bulk of Earth's oceans with water from these sorts of comets,” which come from the so-called Oort cloud in the farthest reaches of the solar system, says Owen. Other planetary scientists tend to agree. “The data are, at the least, discouraging,” says Christopher Chyba of the University of Arizona's Lunar and Planetary Laboratory. “Of course, there are caveats: Other sorts of comets could have furnished our water, billions of years ago.” David Stevenson of the California Institute of Technology agrees: “I think the hypothesis is in trouble, and comets are perhaps less important than we thought in making the oceans.”

The processes responsible for the cloud of primordial material that coalesced into the solar system are thought to have enriched deuterium in some regions and depleted it in others. So planetary scientists are looking for the source of Earth's water in objects likely to have formed in other parts of the cloud. At the meeting, François Robert of the Natural History Museum in Paris announced support for one potential source. He and his colleagues have measured the deuterium-to-hydrogen ratio in carbonaceous chondrites, meteorites thought to date from the early days of the solar system. The ratio varies by an order of magnitude, but the average is close to that in the seas of Earth, suggesting that rocky, chondritelike material—which contains traces of water—could have been the wellspring of the oceans.

Frank, however, thinks the comet theory—his variant of it, at least—is unscathed by the new measurements. “Large comets [such as Hale-Bopp] have nothing to do with the water in our oceans,” says Frank, who believes that the water arrived in swarms of tiny, fluffy objects, whose existence he has inferred from ultraviolet emission observed in Earth's upper atmosphere. These comets, he says, have markedly different compositions from comet Hale-Bopp and its ilk, and perhaps different deuterium-to-hydrogen ratios as well. Then again, no one has yet made a definitive sighting of a tiny comet, let alone measured its composition.

Donald Goldsmith's book, The Hunt for Life on Mars, was recently published by Dutton.

8. BIOCHEMISTRY

# Hijacking a Cell's Chemical Paths to Make New Antibiotics

1. Robert F. Service

The push is on to find new versions of old wonder drugs, as ever more strains of bacteria develop resistance to conventional antibiotics. The trouble is, many antibiotics have a complex molecular architecture that makes them extremely difficult to synthesize, let alone tinker with afterward. For these molecules, medical researchers have largely had to accept what nature gives them: compounds made by antibiotic-producing bacteria and fungi. In this issue of Science, however, a group of biochemists describes a way to hijack the antibiotic-producing chemical pathways of bacteria, exploiting them to produce a wide variety of new compounds.

To construct their natural products, these bacteria rely on an assembly line of about 30 enzymes in which each enzyme hands off its product to the next. Now Chaitan Khosla, a chemical engineer at Stanford University in Palo Alto, California, and his colleagues have managed to interrupt this assembly line partway through, replace the natural intermediate compound with an altered one, and restart the process. As they report on page 367, the new intermediate is taken up by the enzymes and incorporated into the growing chemical structure. By simply changing intermediates, the researchers can get the bugs to construct a variety of antibiotic analogs, many of them well suited to being manipulated further with standard medicinal chemistry techniques.

“It's a nice piece of work,” says Leonard Katz, an antibiotics researcher at Abbott Laboratories in North Chicago, Illinois. “It opens the door to making a bunch of new molecules quickly”—molecules that can then be evaluated as potential new antibiotics, antifungals, or anticancer compounds. Katz is doubtful, however, that the technique could be scaled up for industrial production, as it requires researchers to add synthetic starting molecules that may be time-consuming and expensive to make.

Still, just finding candidate compounds is half the challenge in discovering new drugs. To date, one of the richest sources of promising molecules has proven to be a family of complex natural compounds called polyketides, most of which are made by bacteria and fungi with their involved, assembly-line process for use as defensive chemical weapons. Of the thousands of polyketides discovered thus far, hundreds have already been tapped as pharmaceuticals. That success rate has sent researchers looking for ways to generate new polyketide variants.

By tinkering with the genes for the enzymes that are the “workers” in the assembly-line process, scientists can alter the end product. But “it's still a relatively cumbersome process to make modifications by altering an organism's genes,” says Khosla. What's more, the variety of novel polyketides this strategy can produce is limited, because the organisms must assemble their compounds from molecular building blocks at hand in the cell.

Chemists, of course, can produce a far richer array of building blocks. So Khosla and his colleagues—Stanford postdoc John Jacobsen, Richard Hutchinson of the University of Wisconsin, Madison, and David Cane of Brown University in Providence, Rhode Island—decided to give polyketide-producing bacteria some new building blocks to work with. The researchers started with Streptomyces coelicolor, an organism that is easy to manipulate genetically. They had previously engineered the bacterium to express the entire series of 28 enzymes needed to make the common polyketide antibiotic erythromycin.

Using conventional genetic-engineering techniques, they disabled the third enzyme in the series. That prevents it and enzymes 4 through 6 from working together to take up pairs of a natural, molecular building block—a small, three-carbon chain compound called propionic acid—and from stitching them together into a single six-carbon chain. Enzyme 7 normally takes up this chain and passes it on down the assembly process. But without this six-carbon chain, “the other enzymes don't get what they need to do their thing, and the whole system comes to a grinding halt,” says Khosla.

To reboot the system, the researchers synthesized molecules slightly different from the six-carbon chain that enzyme 7 normally uses as its feedstock and then substituted them for the original. When eight-carbon building blocks were added to the S. coelicolor's fermentation bath, for example, the stand-ins were taken up by enzyme 7 and passed along until a final, new polyketide, possessing two extra carbons, emerged at the end of the assembly line. The result shows that “these enzymes seem to be very tolerant to using new substrates, and that's good news for making novel natural products,” says Khosla.

Enzyme number 7 and its downstream brethren even accepted more radical changes, such as one building block with a six-carbon ring linked to a six-carbon chain, transforming these into a variety of new polyketides. A final processing step at the end to add sugar groups that are normally present on erythromycin turned these products into active antibiotics, says Cane. What's more, the activity of these compounds can be fine-tuned, because they have chemical structures not found in natural polyketides, such as double bonds between adjacent carbon atoms. Medicinal chemists are adept at modifying such structures.

Despite this promise, “the technique still has a long way to go before it's ready to produce molecules for market,” says Katz. He points out that even though S. coelicolor turns out the new products as efficiently as it normally produces erythromycin, other organisms used in commercial manufacture of the drug have been engineered to produce quantities 1000 times larger. And unless the novel polyketides can be produced at this level, it's doubtful the process would be commercially viable, says Katz. He also points out that while erythromycin building blocks are all produced for free as natural metabolites in bacterial cells, the building blocks for the analogs must be specially synthesized, which can be both time-consuming and expensive.

Cane, however, sees no reason why the biochemistry of commercial organisms couldn't be hijacked as well, turning them into factories of new compounds. The need for custom-made building blocks isn't a major obstacle either, Khosla says. He notes that the building blocks he and his colleagues tested took only three or four steps to make, which he calls “within the range of what synthetic chemists are comfortable producing.” A few extra steps could be a small price to pay if the strategy leads to new weapons against antibiotic-resistant bacteria.