Genesis of Eden Diversity Encyclopedia

Get the Genesis of Eden AV-CD by secure internet order >> CLICK_HERE
Windows / Mac Compatible. Includes live video seminars, enchanting renewal songs and a thousand page illustrated codex.

Return to Genesis of Eden?

ns 15jn02

Feed the world? GM crops won't while they're tied to the needs of the rich

THE debate about genetically modified crops has stalled. Greens warn of threats to natural ecosystems, the integrity of organic food, and even our health. The biotech industry has a standard response: "Middle-class Westerners may fuss about such things, but we have a responsibility to feed the starving millions."

This week the shrill voices are being heard again around the corridors of the UN Food and Agriculture OManization in Rome as it hosts the World Food Summit. lt is easy to be cynical of such gatherings. The first, held in 1996, saw governments commit to halving the number of hungry people in the world by 2015, yet do little concrete to achieve that target (see Graph). But clearly there is a shocking problem to solve. Nearly one in seven people continue to go hungry in a world of ever greater plenty.

In a sense, the biotech industry is right to want its technology assessed in a political and economic context. You cannot accurately judge the potential benefits or risks of GM seeds and crops armed only with information about gene flow and plant biochemistry. Just as important are who owns the technology, how it will be traded, and whether it matches the needs and pockets of the world's poorest.

Unfortunately for the industry, such real-world considerations suggest the argument about GM crops feeding the starving millions needs standing on its head. Far from filling empty stomachs, an explosion of GM seeds across the world is likely to spread hunger.

How so? Not because GM crops will trigger environmental catastrophes or epidemics of human illness. These fears have been greatly exaggerated, as the ongoing controversy surrounding GM maize in Mexico reveals (see p 14). What is shocking about that episode is not any threat to maize biodiversity, which is close to negligible, but the sinister lengths the biotech industry seems to be prepared to go to in a bid to stifle debate and railroad the scientific world into rubber-stamping their products. This shifty image will not exactly be helped by the news that some scientists are now attempting to redefine certain types of genetically modified plants as non-GM (see p 13).

Nor are we saying GM crops lack the potential to raise food production worldwide. The present generation of crops may already be able to boost yields in some countries, and one day there may be GM seeds that allow high-yield crops to grow in drought conditions or in infertile badlands. But to conclude from this that high-tech food is the solution to starvation, and that every country must therefore drop its opposition to the technology and open its borders to a free trade in food and seeds, is to completely misunderstand the problem of hunger.

We do not face a world food crisis, and never have. "There is no shortage of food on the planet,' UN secretary general Kofi Annan told the opening ceremony of the food summit this week. "But while some countries produce more than they need to feed their people, others do not, and many of these cannot afford to import enough to make up the gap." He might have added that many countries with enough food cannot feed their starving either. The main reason 800 million people go to bed hungry each night is because they lack the resources to grow or buy their own food.

Poverty is the problem, especially rural poverty. And it is hard to see how a combination of hightech crops and trade liberalised on Western terms will help alleviate it. Poor farmers who struggle to buy even conventional seeds are unlikely to be able to afford the prices charged by the big biotech companies. Nor will they be able to compete with those who can afford the seeds. Farmers in many "food-deficit regions' are already going out of business because their produce is being undercut in local markets by cheap imports from the US, grown on the back of huge subsidies that favour the most mechanised farms. In the short term at least, GM crops will make that problem worse. Overproduction in the West, coupled with a new world order that outlaws trade barriers but allows the subsidies to continue, will see ever more produce dumped in the developing world, further undermining those countries' chances of feeding themselves and increasing the vulnerability of the rural poor.

We are not Luddites or protectionists. Trade and scientific development both have a role in the modern world. But at present, GM crops and the evolving world trade system @hare the same fundamental flaw. Both are harnessed to the needs of the rich world rather than the poor. And until that changes, both are likely to remain part of the problem rather than part of the solution.

Accidental Armageddon What if the world's first nuclear war broke out by mistake?

THE threat of all-out war between India and Pakistan appears to be receding this week. But despite this, fears are growing that the fragile stand-off may yet degenerate into a nuclear exchange. And all through carelessness.

While weapons experts differ widely on how likely that scenario is, many have told Ne-w Scientist that they are extremely concerned about the state of India and Pakistan's nuclear arsenals, which lack many of the safeguards put in place by more estab-lished nuclear powers, They say it is too easy for rogue or misinformed commanders to unleash a nuclear missile or bomber. What's more, a warhead could detonate by accident, making its owner think it had been bombed, and triggering a counterstrike.

India is thought to possess some 35 warheads, and Pakistan between 24 and 48. Both countries claim to keep the warheads "disassembled", with the conventional explosive that initiates the chain reaction stored elsewhere from the nuclear material.

The risk of an accidental detonation depends on how readily the warheads can be reassembled. The Pugwash organisation, which campaigns for nuclear disarmament, quotes Pakistani generals as saying late last year that their warheads can be put together "very quickly", and have no "permissive action links", a security mechanism designed to prevent unauthorised access. The same seems likely to be the case in India.

But aside from fears about unauthorised attacks, simply moving warheads around or loading them onto aircraft or missiles can be risky. Geoff Forden of the Security Studies programme at Massachusetts Institute of Technology notes that the US had to perform many tests on its conventional explosives before it arrived at nuclear warhead designs that would not go off if some of the explosive accidentally ignited. But neither India nor Pakistan has done any such testing. "An accidental explosion would leave little evidence that it was accidental," says Forden. "The government would naturally assume it had been attacked, and retaliate."

American satellites that monitor launches could in theory reassure the aggrieved side that there had been no missile attack. But experts wonder if their advice would be believed, even if it came in time to avert a counterstrike. A joint US-Russian centre for providing such information, which was due to open in May, would add credibility to any reassurance. But it is not yet up and running.

There's no question that accidents do happen. Warheads in the US and the Soviet Union have been engulfed by devastating fires and explosions, though none has ever fully detonated. But M. V. Ramana of Princeton University in New Jersey notes that even if the very sensitive conventional explosives used in Indian and Pakistani warheads blew up without causing a chain reaction, the explosion would contaminate a large area with nuclear material and could cause thousands of cancers.

If there were a full nuclear explosion, it could also trigger events that might lead to a wider war, with China, the US and others rushing to support their allies after the supposed attack. Debora MacKenzie

The great Mexican maize scandal

For environmentalists, the work was proof of the dangers of genetic modification. Transgenes had spread to traditional maize varieties grown in Mexico. Then the journal Nature disowned the research paper, prompting claims that it had given in to a campaign orchestrated by the biotech industry. But why were the authors' strongest critics their own colleagues? Here, we reveal the extraordinary goings-on behind the headlines.

THE saga began in September 2001. The Mexican environment ministry announced that DNA from genetically modified maize had been found in native varieties grown on small farms. The results were, not surprisingly, seized upon by campaigners opposed to GM crops. And in November, they were given more ammunition when the findings were published in the prestigious journal Nature. Then, this April, things took a confusing turn. In an unprecedented move, Nature declared that it regretted publishing the paper and ran two letters that claimed the research was fatally flawed.

The turnaround has generated even more coverage than the original finding. It is the first time Nature has ever disowned a paper in defiance of its authors and referees. Some suspect foul play, claiming that representatives of the biotechnology industry orchestrated a campaign of letters and petitions criticising the original paper. But the paper's critics at the University of California, Berkeley, where the key protagonists in this saga have all worked, may not have needed any outside encouragement. The paper's authors-graduate student David Quist, an environmental scientist, and his professor, Mexican plant biologist Ignacio Chapela, were already hate figures on the Berkeley campus. In 1998, they had campaigned unsuccessfully to prevent the university striking an extraordinary alliance with the Swiss biotech company Novartis. The deal, signed amid student protests and piethrowing, gave Novartis the rights to cherrypick the best plant research for development in exchange for up to $50 million.

But while the protesters see it as compromising academic freedom, many in the Department of Plant and Microbial Biology owe their jobs to the deal with what's now called Syngenta.

Two years later, on the night of 11 October 2000, environmental activists destroyed GM maize being grown at Berkeley by students of Mike Freeling, who is a member of the department. The group told a local paper that they had tested the maize to make sure it was genetically modified.

The angry researchers feared an inside job, and initially pointed the finger at Quist. "Just prior to the vandalism, Quist had requested primers from some of the corn geneticists in my department that might be used to identify transgenics in the field," Steven Lindow, a senior professor in the department, told New Scientist. His colleagues "became concerned, and became even more suspicious after the vandalism", he says.

A fortnight after the crops were destroyed, Lindow spoke to Quist's professor, Chapela, about the allegations. Today Lindow says he quickly accepted that Quist was inno cent. But Quist says that the allegation festers on, and "has led to irrevocable damage to my academic credibility".

At the time of the field trashing, Quist was in the Mexican state of Oaxaca, collecting samples of maize from farmers'fields. When the research based on these samples hit the headlines a year later, it was potentially far more damaging to the careers of biotechnologists at Berkeley than heavy boots in a field at midnight.

'Nature has never said that the paper's conclusions are wrong. We have said that they are not convincing on the basis of the evidence that we have published'

Quist and Chapela first used PCR, the standard DNA amplification technique, to detect the DNA sequences engineered into Bt maize grown in the US. "This is a standard method of detection used by regulatory agencies in Europe and elsewhere," says Quist. It can generate false positives, he agrees. But the pair say that results from their controls show "beyond reasonable doubt" that the sequences are present in a few samples of the native strains from remote regions in Mexico.

The pair then used a related technique called inverse PCR to discover the precise position of the transgenic sequences. This seemed to show that the added DNA had fragmented and scattered throughout the maize genome-the finding that triggered an outcry among scientists.

The authors are still fuelling the dispute. "It suggests that transgenic DNA can move around the genome with a range of unpredictable effects, from disruption of normal functions to modification of expressed products that become toxic agents to the generation of new strains of bacteria and viruses," Quist toldNewScientistthis month. The two critical letters published by Nature in April attacked this second finding. And Quist and Chapela conceded that there were flaws, when, in a letter that Nature published at the same time, they said: "We acknowledge that our critics' assertion of the misidentifying of sequences ... is valid." In less fraught circumstances, a partial retraction of the original paper might have been enough to satisfy both sides. But Nature demanded the authors retract the whole paper, and they refused. So the journal ran its own unprecedented disavowal, in the same issue as the critical letters. This asserted that "in the light of diverse advice received . . . the evidence available is not sufficient to justify the publication of the original paper".

Quist and Chapela point out that, whatever technical failings might have emerged after publication, their paper had been approved by three anonymous referees. It must have had some merit. And when it and the letters of complaint were submitted to three more' referees, two of them specifically noted that none of the comments disproved the conclusion that transgenic corn is growing in Mexico.

"The main finding is not controversial or really being challenged," says Quist. "Neither of the two letters published in Nature, purportedly showing fatal flaws in our paper, ever questioned our main discovery."

Nature has not responded directly to New Scientist's questions about why it would not accept the authors' partial retraction. "Nature has never said that the paper's conclusions are wrong," is all editor Philip Campbell will say. "We have said that they are not convincing on the basis of the evidence that we have published." He denies that a campaign

'Since they seem incapable of admitting their mistakes, they are raising non-scientific issues like the Novartis agreement, vendettas and global conspiracies'

against Quist and Chapela influenced his decision to demand a retraction of their paper-and to disown it when they refused.

But a campaign there certainly was. Demands that the paper be retracted appeared on Internet biotech forums the day it was published, and continued with mounting vehemence. Yet two of the first, most persistent and apparently scientifically qualified complainants on the Net, "Mary Murphy" and "Andura Smetacek", appear not to be real people. A British anti-GM campaigner, Jonathan Matthews of the Norfolk Genetic Information Network, claims to have tracked their electronic personas to the offices and computer equipment of the Bivings Group in Washington DC, a PR company that has Monsanto as one of its clients. Bivings initially denied everything but has since admitted that one of the emails came from a Bivings' employee or client.

But what has raised most eyebrows is the identity of the scientists whose two letters attacking the paper appeared in Nature. "The antagonists signing the letters are all connected directly with [Berkeley's] local political scandal," says Chapela.

One was co-written by Freeling and Nick Kaplinsky, who is also a senior figure at the same department at Berkeley. The other was by Matthew Metz, a former Berkeley microbiologist who was a vocal supporter of the Novartis alliance, and johannes Futterer, a young Swiss researcher whose boss, Wilhelm Gruissem, was at Berkeley four years ago and was widely regarded as "the man who brought Novartis to Berkeley". Quist and Chapela believe the animosities created by the furore over the deal, and inflamed by the crop trashing, must be an element in the row over their paper. Kaplinsky denies this. "This issue is strictly about science. Quist and Chapela published bad science and should have done the honourable thing-retract their paper and apologise." But Kaplinsky doesn't stop there. "Since they seem incapable of admitting their mistakes, they are raising non-scientific issues like the Novartis agreement with our department, vendettas, global conspiracies.

Anything so they can avoid talking about the fact that they published artefactual data and then misinterpreted it."

Campbell says he wasn't aware of the allegations surrounding the crop trashing incident when he accepted the letters. But he says it would not have influenced his decision to publish. Neither Nature nor Campbell are poodles of the biotech industry. Campbell himself wrote a hostile editorial about the Berkeley-Novartis deal. But Quist still insists that it was political pressure that brought about the journal's actions.

Whoever is right, the row reveals an alarming breakdown in scientific discourse. In the aftermath of the affair, Campbell wrote that it must have been Murphy's law that ensured the journal's embarrassing climbdown "was in relation to one of the most hotly debated technologies of our time". Others see it as more than an accident. They fear that the affair has put the system of peer review to the test, and found it wanting.

The spectre of unseen actors manipulating events is particularly worrying. In its disavowal, Nature asked its readers to make up their own minds about the science behind the row. But it failed to alert them to the private rows behind the public letters. Nor did it reveal the identities and affiliations of the five referees who broadly supported the original paper, or the sixth who appears to have persuaded Nature to make a retraction.

Also out of sight are the individuals behind "Mary Murphy" and "Andura Smetacek", not to mention the people who trashed Freeling's field two years ago. Strange what dark shadows are thrown up by the harsh glare of publicity. Fred Pearce

Is it worth worrying about?

WHETHER or not Chapela and Quist's work stands up to scrutiny (see main story), there's no denying that transgenes have spread to traditional maize varieties cultivated in Mexico.

Last month, researchers told a conference in the Netherlands that two separate teams have found transgenic DNA in around 10 per cent of crop plants sampled in Oaxaca province. "it is confirmed. There is no doubt about it," says Jorge Sober6n of the National Biodiversity Commission in Mexico, who described it as "the world's worst case of GM contamination". Contrary to many reports, however, no studies have found any evidence of transgenes in wild maize.

The genes spread despite a ban on the planting of GM maize that Mexico imposed in 1998 specifically to prevent such contamination. Maize was bred from wild grasses in Mexico 6500 years ago, and a rich selection of strains is grown by small farmers. Despite the precautions, transgenic Bt maize imported from the US for food appears either to have been deliberately planted or spilt in fields, and the transgenes are now probably also spreading by cross-pollination. So does it matter? Some think the effect will be to wipe out diversity. GM contamination "may destroy the genetic integrity of the diverse varieties in these centres that are humanity's insurance against future threats from disease, climate change and biotechnological calamities", says Patrick Mulvany of the British-based Intermediate Technology Development Group. "Governments should declare an immediate moratorium on t@'e release of GM crop seeds and grains in centres of crop diversity."

There is no doubt that old varieties and wild relatives of crop plants are a valuable resource for breeders. But they're already disappearing fast, and the main threat is not GM pollution but the abandonment of traditional varieties in favour of modern hybrids.

Mexico's problem isn't likely to go away. Bt maize contains a gene for a bacterial toxin that kills some pests. This gives it an advantage over other strains, making it more likely to spread further even if farmers don't knowingly choose to plant these seeds.

But this won't necessarily spell doom, according to the International Maize and Wheat Improvement Center (CIMMyr) in Mexico. lt says diversity could actually increase as a result. The Bt gene won't wipe out other genes, but will make the overall mix that little bit richer. And if plant scientists find a desirable trait in a contaminated variety, they can easily breed plants that contain the desired trait but lack the Bt gene.

Nor does the contamination pose any particular threat to the two types of wild maize found in Mexico, teosinte and Tripsacum. Teosinte only occasionally interbreeds with cultivated varieties, while hybrids between Tripsacum and maize are even rarer.

lt seems unlikely that the contamination of Mexican maize with the Bt gene will have any serious consequences. But it is deeply disturbing that extensive gene flow took place despite Mexico's efforts to prevent it. CIMMYR warns that other kinds of genes could spell trouble. Those designed t6 produce industrial products such as medicines, for example, might make maize unsafe to eat, and such contamination would be very difficult to deal with. Next time it could be worse.

Such fears don't appear to trouble the industry. "it is better to acknowledge that a minimum of cross-pollination cannot be avoided, and not to panic," said Guy Poppy of the British biotech association CropGen, after Chapela and Quist's paper fi rst appeared.

There are ways of preventing genetic pollution, such as "Terminator" technology (see New Scientist, 28 October 2000, p 4). Perhaps now is the time to start using them.

The road to extinction Even the smallest open spaces spell doom for Amazonian wildlife

ANIMALS and plants in the Amazon are being hit far harder by people encroaching into the jungle than anyone thought. The biggest investigation to date has concluded that Amazonian wildlife is agoraphobic, avoiding even the smallest clearing in the vast sea of rainforest, and that fragmentation of the forest is having a massive impact on native species. In the review, that analyses the findings of 340 separate studies, researchers found that fragmentation increased local extinction rates and altered species richness and abundance. It also disrupted ecological processes and increased the risk of fire.

Fragmentation has a much larger impact on tropical rainforests than on other ecosystems because natural clearings and fires are rare. And the immense size of the Amazon means there are many extremely specialised animal and plant species that are not adapted to life on the edge.

"It was a surprise that even small unpaved roads can have an impact," says William Laurance of the Smithsonian Tropical Research Institute in Balboa, Panama, who led the research team. For instance, some birds that live in the lower canopy or on the ground won't go near the edge of the forest, let alone cross a narrow road. "There are fewer species 70 and even 100 metres from the edge of a road."

Logging, roads, pipelines and fire continue to chop the Amazon into more and more fragments. "Forest edge" has increased at the rate of 19,000 kilometres per year since 1992, estimates Mark Cochrane from Michigan State University in East Lansing. And the rate of deforestation may now be worse than it was in the 1970s. Also, despite Brazil's efforts to educate farmers and impose controls, slash and burn continues to be a common agricultural practice. In addition, over 90 per cent of accidental fires start around forest edges, he says. "Tropical forests generally don't burn unless they are chopped into smaller fragments."

Laurance's group report in this month's issue of Conservation Biology that species living within forest fragments smaller than 100 hectares are doomed. While they could not determine a minimum critical size for nature reserves, the researchers believe they need to cover millions of hectares to preserve the low population densities of some species. And because the composition of species varies drastically from area to area, a large number of reserves, linked by wide corridors of primary forest, are also essential to maintain healthy levels of biodiversity.

That could prove all but impossible, as fragmentation will be on an unprecedented scale by 2020, Laurance says. Brazil has already given the go-ahead for a $40 billion "Advance Brazil" programme for new highways, hydroelectric schemes, gas pipelines and river-channelling projects in the Amazon rainforest. Even the Porto Velho-Manaus and Cuiaba-Santarem superhighways now being built will spark profound changes, as they'll bring loggers, miners, hunters and settlers into pristine areas. Stephen Leahy

GM Therapy in a Glass

A COLONY OF GENETICALLY MODIFIED BACTERIA LIVING AND BREEDING IN YOUR GUT COULD BE JUST WHAT THE DOCTOR ORDERED. BUT IS IT WORTH THE RISK, ASKS GAIL VINES

IT'S such an obvious idea, it's surprising no one thought of it before. Everybody knows that microorganisms can be genetically altered to pump out virtually any human protein to order. And it's also clear that lots of diseases could be treated with proteins, if only it were possible to get the right dose to the right place at the right time. Up to now, that's been the stumbling block. But perhaps there's a way. Just swallow some genetically altered bugs and let them make a home in your gut. Hey presto: a cheap and steady supply of home-made medication-for life, perhaps. It's a medical revolution already waiting in the wings. Researchers in Belgium and the Netherlands are poised for the first trials in humans. If the tests get the go-ahead, it will be the first step towards a radically new kind of therapy: using GM bacteria to deliver therapeutic proteins just where they're needed. Sounds promising. But before anyone swallows a single GM bug, the researchers will have to prove that it's safe. And not just for the patients who'll be asked to turn their intestines into GM bioreactors. Once the modified bacteria are released into the environment, there's no recalling them, and no way of knowing if they'll stay benign. Is that a risk we're prepared to take?

Molecular biologist Lothar Steidler of the Flanders Interuniversity Institute for Biotechnology in Belgium thinks it should be. He heads the team that's pioneering the new approach with a GM bacterium designed to cure inflammatory bowel disease. IBD embraces a range of debilitating and sometimes fatal chronic conditions, including Crohn's disease and ulcerative colitis. It's a major health problem, particularly in the West, where as many as 1 in 1000 people suffers from one form or another.

Bowel disease is just the start. If the new strategy works, it could be deployed against a host of other diseases. "The panorama of applications is huge," says Steidler. He's reluctant to be more specific for fear that patients will ring his lab hoping for miracle cures. But in theory, any protein drug that can be absorbed through the gut wall into the bloodstream could be delivered in this way. One obvious candidate is insulin, which can't be taken orally because the stomach digests it before it can be absorbed. First, though, Steidler wants to see whether his bacteria will behave as expected in people. The first human trials, which are awaiting the nod from the Dutch government, will take place in Amsterdam. The plan is to give a dozen or so volunteers a strain of Lactococcus engineered to manufacture a human protein called interleukin-10. IL-10 is a potent "immune modulator", capable of dampening down excessive immune reactions. The hope is that a steady trickle of IL-10, delivered straight to the gut wall by the GM bacteria, is just what is needed to suppress the abnormal immune response that gastroenterologists believe underlies IBD. Steidler's group originally set out to use bacteria as a cheap way to produce IL-IO for other researchers. In the process, they stumbled across Lactococcus. 'Suddenly it struck us: perhaps we could use this bacterium directly as a therapy,' he says. "The conventional approach is simply not economic. To produce and purify enough IL-IO to treat one patient costs something like f 10,000 per annum, and a patient might need life-long therapy. By contrast, our approach seems very cost effective." Therapies for IBD already exist, and the most effective is probably repeated treatment with corticosteroid drugs, which also suppress the immune response. But because steroids bludgeon the entire immune system, they can have undesirable long-term side effects. The promise of the GM bug approach is precision engineering. All you have to do is drink a liquid suspension of the bacteria once a day and they colonise your bowel, releasing the requisite protein all the while.

It's an ingenious solution. All the same, the prospect of GM organisms setting up home in your colon isn't particularly enticing. By way of reassurance, Steidler first points out that the notion of eating therapeutic bacteria is not new. As early as 1907, the Russian microbiologist Elie Metchnikoff claimed that such "probiotics" have many health-giving properties and could even prolong life. Modern research has largely substantiated his arguments (New Scientist, 17 November 2001, p 40), and probiotic foods such as Actimel and Yaktilt are now mainstream.

Steidler's starting point, Lactococcus lactis, is a "food-grade" bacterium routinely grown in huge vats by cheese and yogurt manufacturers. Lactococcus is part of your normal intestinal flora and, along with other bacterial groups such as lactobacilli and bifidobacteria, is the not-so-secret ingredient in a vast range of fermented foods including yogurt. Every time you down a "live" yogurt, you are bolstering the indigenous Lactococcus population of your intestines.

These microbial allies contribute to a healthy gut by binding to its lining, making it harder for hostile microbes to attack. What's more, the friendly organisms also obligingly release compounds that we now know are potent anti-cancer agents.

But even if undoctored Lactococci are so beneficial, will the same be true if they are genetically engineered? There are obvious safety concerns. What might be the outcome of person-to-person transmission of such organisms? And what might happen if the gene for IL-IO escaped from the modified Lactococci and spread into some other bug?

Sue Mayer of the British campaign group GeneWatch agrees there's a need for caution. "We know so little about the behaviour of microorganisms generally," she says, "that it is particularly difficult to know whether genetically modified strains are going to pose any particular threat." Doug Parr, chief scientific adviser to Greenpeace UK, adds: "If this leads to the deliberate release of a GM organism into the environment then there are potentially big issues." And the campaign groups are not alone in their concerns. "There is an awful lot that goes on in our guts that we're still totally in the dark about," says geneticist Michael Antoniou of Middlesex University. The prospect of genes for powerful immune-system modulators breaking loose and spreading out of control is frightening indeed.

The trouble is, it's impossible to predict exactly what risks are associated with a release of this kind. Up to now, the only application that has involved the deliberate release of GM microorganisms has been in bioremediation, using modified soil bacteria to digest pollution. The only existing medical application that comes close is gene therapy-ferrying missing genes into patients with inherited diseases such as cystic fibrosis using modified viruses, for instance.

'THERE ARE OBVIOUS SAFETY CONCERNS. WHAT WOULD HAPPEN IN THE EVENT OF PERSON-TO PERSON TRANSMISSION?'

Steidler's approach is distinctly different from either of these. His team's proposed trials would be the first time a GM bacterium with an inserted human gene was released into a human body, and potentially into the environment at large. Aware of the problems this might cause, Steidler and his colleagues have done all they can to ensure that their bacterium poses no threat. For starters, the human IL-10 gene has been integrated into the bacterial genome, rather than being left floating on a separate "plasmid" loop of DNA within the body of the bacterium. Steidler say this should eliminate the risk of the gene jumping to another species. Plasmids are notoriously promiscuous, but genes within the main genome rarely play the field.

In addition, the modified organism is an "attenuated strain" that can't survive without a particular nutrient included in the growth medium. Without a steady supply the bugs die within a couple of hours. Steidler won't say what the nutrient is except that it's an ordinary food supplement that patients would swallow along with the bacteria.

As a result, any patient who wanted to evict the GM bacteria could simply stop taking the supplement, and the bacteria would quickly die. What's more, says Steidler, the supplement has been fine-tuned so that very little reaches the end of the colon. This precaution ensures that the engineered bacteria die soon after being expelled in faeces, and so can't spread to the environment or other people.

Of course, there is always the danger that the modified bacterium could acquire a gene that allows it to live without the supplement. Bacteria are remarkably good at swapping genes with their fellows, even across species barriers. But such an event is generally regarded as so unlikely as to be negligible.

'BUT IT MIGHT BE WORTH THE RISK ' TRIALS IN MICE SHOW THAT THERAPY WITH GM BACTERIA REALLY MAKES A DIFFERENCE'

And in any case, it may well be worth the risk. Extensive trials in mice suggest that the therapy really makes a difference. Almost two years ago, Steidler's group published encouraging results. Mice with a condition similar to human inflammatory bowel disease showed up to a 50 per cent reduction in gut inflammation, making the approach just as effective as steroid treatment but with potentially fewer side effects (Science, vol 289, p 1352).

Yet worryingly, perhaps, the researchers still have no clear idea how the GM bacteria exert their beneficial influence. They might simply attach themselves to the gut wall and secrete IL-10 which influences the target immune cells. But another possibility is that the bacteria are actually taken up by specialised immune cells called "M cells" in the gut lining. In this case, the bacteria would be incorporated into the immune system itself, releasing IL-10 from the M cell. It might make a big difference which mechanism is at work, so shouldn't we find the answer before we start any human trials?

Steidler says that his team is setting up a series of research projects to find out. All the same, he's convinced that human studies should start immediately. The first step, he says, is a small-scale trial to see whether the experimental system is safe and effective. The "phase I" trial in humans is being planned by Steidler's medical collaborators at the Academic Medical Center in Amsterdam. The researchers want to make sure that neither the bacterium nor the IL-10 has unexpected side effects. They also want to see whether the bacteria colonise the right part of the colon. Gastroenterologist Henre Braat of the AM@ says that if they get the go-ahead, they'll feed the GM bug to 12 patients with Crohn's disease.

During the trial, the patients will' be admitted to hospital, but won't have to stay in a special environment "Our main concern is to ensure they use a chemical toilet," Braat says. All their faeces will be collected and destroyed so no modified bacteria escape.

On 21 May, the Dutch scientific advisory committee on genetic modification gave "draft" approval to the trial. The publication of the draft licence in newspapers and on a government website will be followed by a four-week consultation period "in which there is time to answer any objections the public might have", says Birgit Loos, the head of the Dutch civil service office in charge of GM organisms. If there are no objections, then the licence will be finalised and there will be a further six-week period for public objections. If all disagreements are resolved within that time then the trial will go ahead.

The outcome of the public consultation period is hard to predict. Loos says she expects the licence to stand, based on past experience: medical biotechnology applications always go smoothly, she points out. But this isn't any old medical biotechnology trial. It is officially classed as an environmental release of a GM organism, and setting GM organisms loose is widely unpopular. We can only wait and see how this one plays out. What is clear, though, is that the Dutch trial is just the tip of the iceberg. Steidler's group is already looking for other potentially therapeutic proteins that could be manufactured by obliging bacteria. For instance, "trefoil" peptides might do for acute immune reactions what IL1 0 promises to do for the chronic sort.

And we might one day send drugs to different regions of the digestive tract. just as Lactococcus settles in the colon, perhaps it'll be possible to engineer a bacterium to deliver drugs to the stomach or small intestine.

Many research groups are now convinced that one way or another, GM bugs-perhaps in the form of live foods such as yogurt-will eventually be used as living drugs couriers. Steidler has put his money on Lactococcus as the vehicle, not least because it can easily be engineered to churn out proteins continuously and in large quantities. Meanwhile a French team led by Monique Alric of the University of Auvergne in Clermont-Ferrand is backing gut-dwelling yeasts.

'GM BUGS IN YOUR GUT COULD EVEN MOP UP TOXIC WASTE PRODUCTS SUCH AS URIC ACID, FREEING KIDNEY PATIENTS FROM DIALYSIS'

Yeasts are bigger, and in some ways trickier to work with than bacteria. In yeast, human genes usually need to be switched on by some other compound, and often yield tiny amounts of the product. On the other hand, because yeasts are much more closely related to mammals than bacteria are, they can make a wider range of human proteins.

So far the French team have concentrated on creating experimental yeasts and testing them in a Ipb model of the human gut. The group already has yeasts capable of secreting compounds such as a plant protein called cytochrome P450, and others that can biochemically convert one compound into another. At the moment the team is negotiating contracts with commercial firms to insert genes for therapeutic proteins into yeast. The French approach might also help assess the level of risk before human trials begin. 'One advantage of our system is the possibility to see if gene transfer takes place between the yeast and different populations of gut bacteria," says Alric.

Another intriguing possibility is using GM bugs to mop up toxic waste products such as uric acid and creatinine, which build up in the blood of kidney patients and currently have to be removed by dialysis. Researchers at the Artificial Cells and Organs Research Center at McGill University in Quebec have already reported that artificial cells containing a GM strain of the gut bacterium E. coli can lower levels of uric acid and creatinine in rats.

It might be risky,.but i ' f the new approach makes curing serious illnesses such as Crohn's disease and kidney failure as easy as making yogurt, surely it's worth a shot.

Sneaky Sex

SEX is a risky business in the animal kingdom, especially for the guys. Take the dangerous love life of the South American knife fish. Males attract a mate by emitting electric signals over a wide range of frequencies. Unfortunately for them, their low "notes" are such an effective beacon to predators, including electric eels, that by the end of the mating season there are almost no male knife fish left.

But what else is a felia to do, when females seem to go for the gaudiest, loudest males? American song sparrows, for example, are hard to miss when courting. The male sings loudly from a branch out in the open, easily seen and heard by hungry hawks as well as female sparrows. And it's the same for countless other species. If the coded flashes of a male firefly are spotted by a female of a different species, the hapless Romeo is likely to end up as dinner rather than a mate. In Panama, predatory bats home in on the night-time love calls of male tungara frogs. And a male cricket's song can earn him a mate or attract a parasitoid fly that lays its eggs on the suitor, sealing his doom. Why do males have to put on such a show when it puts them at obvious risk? The question has been hotly debated by the finest minds in biology from Darwin onwards. Perhaps showy males are proving that they are fitter and healthier than drab ones, which is currently the most popular hypothesis. Or it may be outlandish female tastes that are driving prospective mates to evolve flamboyant sexual displays-the theory of runaway sexual selection. But neither of these theories tells the whole story because it turns out that females aren't always attracted to risk-takers, and on closer scrutiny, even high-profile males can be more cautious than it might appear. Even the seemingly suicidal knife fish hedges his bets. Philip Stoddard of Florida International University in Miami has found that male knife fish emit high-frequency electric signals throughout the day, but only crank up their low-frequency output during the evening, when females spawn. "If they don't use low frequencies, they're not gonna get the babes," says Stoddard. But by emitting the full range of signals only in the evening, they minimise the risk of being eaten and maximise the chances of passing on their genes to the next generation.

Sexual balancing act

And many other animals are even more subtle. Lacewings serenade their mates by vibrating the plants they rest on, rather than producing a song that could be heard by predators. The feathers of the male Guianan cock-of-the-rock, which are brilliant orange in sunshine, fade into invisibility in forest shade, where it lurks when not showing off to females. And when hungry bats are on the prowl, male tree frogs utter shorter calls or sing near loud waterfalls that mask the sound. Even more surprisingly, it turns out that under certain circumstances females may actually find discreet males more sexy than showy ones. This is the case for the Trinidadian guppy, a tiny fish that has taught scientists more about the sexual balancing act than any other animal. "Guppies specialise in blue, red and ultraviolet-colours that their predators can't see," says John Endler of the University of Califomia at Santa Barbara. They court early in the morning and late in the afternoon, when their predators are least active, and when the low light shows up their colour spots to best effect for appreciative females. But Endler has also shown that where predators are common, male guppies have evolved drab colours compared to the bright spots they flaunt where predators are few.

One explanation is simply that brightly coloured males are more likely to be eaten, leaving the dull ones to reproduce. But drab males may prevail under these circumstances because females like them that way. Laboratory experiments show that females are more likely to mate with a drab male if they know a predator is nearby.

Danger may force them to be less picky. 'If a female chooses a mate at random it's quicker-she spends less time around conspicuous males, whose proximity can make her more vulnerable to predation," says Endler. Alternatively, female guppies may find dull colours more appealing when the threat from predators looms large.

Either mechanism could explain the patterns of guppy evolution Endler has seen in the lab and in the wild. Whichever it is, a female's change in preference is likely to be innate rather than learned. 'teaming is slow and might be dangerous," says Endler, 'so a genetically programmed reaction to predators would have higher fitness and is likely to spread through the population." This isn't the only instance of females changing their preferences depending on the situation. Crickets do it too. Males produce their mating trills by rubbing two wing parts together, like a bow against a violin string. They do this from the relative safety of a crack in the ground. All things being equal, females will always choose males with long trills over ones with short trills, but Ann Hedrick of the University of California at Davis has found that females factor in the risks of predation when picking a mate.

Hedrick played recordings of long or short trills in an experimental mating arena that contained both cover and open space. Then she waited to see which call drew the female. The outcome depended on where the call was coming from. "I offered females a choice between a short call with cover in front of it versus a long call in the open,' says Hedrick. "When you put them in that conflict situation, where they hear the call that they like but it feels unsafe to go that way, many of the females instead go towards the short call through cover." With no cover on offer, the same females throw caution to the wind and always choose the long call. It's not clear why female crickets are attracted to long trills. They may be an indicator of quality-an example of the "handicap principle" whereby sexual signals must cost the animal something to prove their worth. "We know that trilling takes a lot of energy," says Hedrick. Alternatively, males could have been forced to evolve ever-longer trills simply because females like them. This process of runaway sexual selection results in a trait becoming more and more exaggerated until its cost to a male's survival outweighs the benefits of attraction. Evidence that male crickets take it to the limit comes from Hedrick's discovery that those with the longest trills are also the most cautioushunkering down and hiding in their cracks for longer when predators are about. 'There's more than one way of being an effective male cricket," concludes Hedrick.

Another way to mate effectively without drawing attention to yourself is to exploit the bold display of a less careful guy. Male wax moths have been known to hang about silently near a singing rival, then waylay females who come in search of the vocalist. Some male crickets also loiter near competitors and seem to be using the same strategy.

Sneaky ploys are found among the males of many other species, including guppies. Here, the cheats bypass the whole courtship ritual, which entails quivering in front of a prospective mate to show off your colours. They choose instead to force themselves on unresponsive females. The strategy is more common when the risk of predation is high, when females tend to lose interest in sex.

In this case there's no indication that females actually find the cheats attractive. But bluegill sunfish may. Here males come in two kinds: big guys who defend territories and display to females, and little guys who look similar to females. The smaller males don't court females directly. Instead, they wait near a big male's territory, and when a female swims in to lay her eggs, the little guy darts in and releases some sperm.

Many biologists see the bluegill sunfish as a classic example of a sneaky mating strategy, but not Amotz Zahavi of Tel-Aviv University. He's the originator of the handicap principle theory, and believes there's no sneaking going on here, just honest signalling of intention on the part of all three players. He argues that females prefer territories where their eggs will be fertilised by both kinds of males-so a big male should welcome a small male, because the two of them together will get more mates than either would alone.

Zahavi argues that any communicationsexual or otherwise-must be associated with some cost, or handicap, to prove its honesty. "Where two individuals collaborate, the signal must be loaded with a handicap, there is no other way," he says. Zahavi acknowledges that when the cost of a signal becomes too high, it will disappear-or become covert, like the love-vibrations of the lacewing. But he believes that nature demands honesty in courtship, and that behaviour now seen by many biologists as 'cheating" will ultimately be explained in terms of straightforward signalling.

The handicap principle is the most popular explanation around for the evolution of showy males, but the discovery that it isn't the only way to woo has left some biologists questioning a theory that puts honesty at its core. "Zahavi thinks the handicap principle is everything," says Endler. "His argument isn't wrong, it's just incomplete." Endler argues that any kind of mating strategy can evolve provided it works-that is, it will increase an individual's chances of producing offspring that will adopt the same strategy.

Is honesty irrelevant? Is it anthropomorphic to describe evolution in terms of honesty? Can a single principle really explain the bemusing array of sexual behaviour found in the animal world? The debate will continue, as surely as fireflies glow on June evenings. But it's worth remembering, the next time you see those ethereal glimmers across a summer meadow, that there's more to courtship than meets the eye.