Washington Post, January 20, 2004

GENETICALLY MODIFIED ORGANISMS NOT EASILY CONTAINED

National Research Council Panel Urges More Work to Protect Against Contamination of Food Supply

By Justin Gillis

Techniques for confining genetically engineered salmon, corn and other organisms are still in their infancy, and far more work needs to be done to make sure the new products do not taint the food supply or wipe out important species, a National Research Council panel said today.

To date, most attempts to control potentially hazardous, gene-altered species that are grown outdoors have involved establishing physical barriers, like rows of trees, or altering planting times to make sure crops cannot cross-breed with related plants nearby. But those techniques have proven susceptible to human error, and researchers have long recognized that physical methods are likely to become even less useful as gene-altered insects and other animals begin to emerge from the nation's laboratories.

Scores of altered organisms are under development, offering numerous potential benefits -- and many theoretical perils. While eager to reap the benefits, many scientists are worried that gene-altered crops might breed with wild relatives to produce super-weeds, for instance, or that genetically engineered salmon or honeybees might kill off their wild relatives by out-competing them for food.

Scientists have been studying newer technologies that might impose biological limits on the movement of genetically engineered species or the spread of their genes. But the most promising methods of "bioconfinement" are still in the early research stages, and no available method offers complete assurance that new products deemed especially hazardous can be kept under control, the panel said in a 219-page report commissioned by the Agriculture Department, which is charged with regulating many aspects of genetic engineering.

"What they seem to suggest is the science for creating risky organisms exists, but we don't have the methods for safely confining them yet," said Gregory Jaffe, director of biotechnology programs at the Center for Science in the Public Interest in Washington. "The sad conclusion from the report is that there really aren't any viable bioconfinement methods that could be adopted commercially without significant additional research and testing." Jaffe's organization is a consumer group that supports genetic engineering in principle but has often criticized federal oversight of it.

The National Research Council panel emphasized that many types of gene-altered organisms pose little or no theoretical risk, and control techniques won't be needed. For the minority of organisms that do pose risks, the panel recommended that companies and laboratories adopt an "integrated confinement system" that includes at least two distinct techniques. The plans should be overseen by regulators in Washington and should factor in the likelihood of human error, the panel said, adding that confinement had sometimes seemed to be an "afterthought" in genetic-engineering research.

If widely adopted, the recommendations would impose new costs and burdens on the U.S. biotechnology industry. While emphasizing its commitment to safety, the industry has generally opposed elaborate control methods for gene-altered organisms, saying the risks have been exaggerated and the potential benefits under-appreciated.

Val Giddings, vice president of agriculture at the Biotechnology Industry Organization, a Washington trade group, noted that gene- altered organisms have been used inside laboratories for decades with an excellent safety record, and altered crops have been widely planted since the mid-1990s. "We have hundreds of millions of tons of this stuff being grown around the world for years, and eaten by millions of people, with literally not a headache or a sniffle yet," he said.

Anne R. Kapuscinski, a member of the panel and a fish biologist at the University of Minnesota, said at a briefing in Washington today that the techniques of genetic engineering offer "enormous potential for modern agriculture" and for solving other problems. But as scientists design ever-more-exotic organisms -- ranging from corn that produces pharmaceuticals in its kernels to fish that grow 10 or 20 times faster than normal -- the risk will rise that altered genes could spread to new species or unwanted locales, threatening the ecology or the food supply, the report said.

That nearly happened in 2002, when human error allowed corn designed to produce a pig vaccine to spread too widely in fields in Iowa and Nebraska. Expensive, last-minute intervention by the Agriculture Department kept the product out of food, and the department has since been tightening its regulations. Some advocates of genetic engineering have charged that regulation has already become excessive and threatens to choke off one of the nation's most promising new industries, while environmental and some consumer groups assert that the government hasn't cracked down enough.

The new report was commissioned before the corn incident, but has taken on added importance in light of that near-miss. The National Research Council is the research arm of the National Academy of Sciences, the National Academy of Engineering and the Institute of Medicine, the nation's three most prestigious scientific advisory bodies, and its reports generally carry weight with all political factions in Washington.

Many scientists have said that confinement, or lack thereof, is proving to be the Achilles' heel of genetic engineering. The gene- altered food crops commercialized to date -- the most important are soybeans, corn and canola -- have turned up repeatedly in unexpected places, including overseas shipments meant for markets that won't accept gene-altered ingredients.

Some newer organisms under development promise to be even harder to control. Plants, after all, are stuck in place with roots in the ground, but gene-altered animals will be capable of moving on their own.

The various bioconfinement techniques available today all suffer from problems that undermine their reliability, the report said. It noted that scientists are working on potentially better techniques. For instance, a plant could be engineered so that its flowers always die before spreading pollen, or an animal could be made dependent on some man-made substance so that it would die if it escaped. But research on these methods is just beginning and long years of work lie ahead, the report said.

As a case study of the difficulties, the report offered the example of a fast-growing salmon under development by Aqua Bounty Technologies Inc. of Waltham, Mass. The gene-altered salmon reaches market size in half the usual time, requiring less feed. Aqua Bounty wants to sell the fish for use in ocean pens along the East Coast, where other farm- raised salmon are grown. The company has acknowledged that some fish will inevitably escape, but has said they will be so dependent on food supplied by humans that they are likely to die in the open ocean.

Environmentalists are worried that the fish, which they have dubbed "Frankensalmon," would not die, but instead would wipe out dwindling stocks of wild Atlantic salmon by competing with them for food and, among males, competing for access to wild females. To meet these concerns, Aqua Bounty plans to sell only sterile, female fish. But the new report said the methods for sterilizing the fish are not entirely reliable, and it urged that the Aqua Bounty fish be tested individually for sterility or grown only in tanks on land -- expensive methods that most fish-farming companies are likely to resist.

Joseph McGonigle, a vice president at Aqua Bounty Technologies, said his company was still evaluating its production techniques and the report was premature in drawing conclusions about how reliable they would be.

"All of this is really just sound and fury," McGonigle said. "Nobody has any evidence, and it's not going to be there until we put it on the table. We're certainly aware of the risks."

2004 The Washington Post Company