Druggenes could enter food chain

BYPhilip Cohen; San Francisco

Loopholes in US regulations raise fears that food crops will be contaminatedwith pharmaceuticals

THE rules the US government is proposing for field tests of crops that have beengenetically modified to produce pharmaceutical products are not strict enough toprevent the contamination of food crops, experts have told New Scientist.

They say the proposed rules are based on flawed science, that there areloopholes allowing them to be bypassed, and that companies do not even have todisclose what genes have been added. And they warn of severe environmentalconsequences if a drug-laced plant were to breed with other crops or wildrelatives.

Biotech companies plan to produce a vast range of products, from drugs tovaccines, in plants. "These plants have the potential for more benefit thanany other agribiotech product," says Jane Rissler of the Union of ConcernedScientists in Washington DC. "But to realise those benefits we have to bevery careful about the risks." The US Department of Agriculture's (USDA)plans don't come close, she says. What scares Rissler and others is that therecould be a rerun of the Starlink debacle, in which GM corn strictly not intendedfor human consumption ended up on grocery shelves. If any contamination involveda crop producing a potent drug, the consequences could be far more serious, shesays.

The proposed rules require the "pharmed" plants to be separated fromother crops in time as well as space. For example, pharmed maize must be grownat least 400 metres away from other maize. It must also be planted two weeksbefore or after nearby crops, so that it isn't fertile at the same time. Similarregulations have been outlined for other plants that have been engineered tomake drugs, including barley, corn, rice and sugar cane. But when it comes tokeeping harvested products separate, the rules are vague, talking only of"adequate identification, packaging and segregation".

Companies that violate these procedures can be fined $250,000, andindividuals could face jail sentences of up to five years. James White, theUSDA's branch chief for biotech evaluations, is confident the rules will dotheir job:"The chance of gene flow is essentially zero."

"These rules are more stringent than prior recommendations, and I applaudthat," says Norman Ellstrand of the University of California, Riverside,who sat on a National Academy of Sciences committee that reviewed theregulations for GM crops. "But there are big holes in the system."

The NAS report points out that some of the USDA's rules have no clear scientificrationale. For instance, the isolation distance for corn is simply double the200 metres it recommends for the production of GM seeds. The assumption is thatthis spacing will reduce contamination to 0.1 per cent, but there is no evidencethat the contamination risk drops off with this increase in distance. Only lastweek, Australian researchers reported that pollen from oilseed rape hadcontaminated fields up to 3 kilometres away, and that there was no obviousdrop-off with distance (Science, vol 296, p 2386).

Another serious concern is that the USDA focuses on the intended use of a cropproduct and ignores its other possible impacts. For instance, the Texas-basedcompany Prodigene applied to grow maize that produces a chicken-egg proteincalled avidin, which is known to kill or harm 26 species of insects. But becauseavidin is not classed as a drug, the crop doesn't come under the pharmingregulations. Nor did the USDA look at the maize's environmental impact becausethe crop wasn't being grown in order to kill insects. "If they had used thesame protein as an insecticide, they would have called in the EnvironmentalProtection Agency to regulate it," says Ellstrand.

While avidin's properties are well known, that is not the case with every drugthat might end up being grown in crops. Ellstrand and his colleagues weredisturbed to discover that the publicly available descriptions of genes splicedinto some plants are incredibly vague.

White says the USDA will start posting fact sheets on genes in transgenic plantslater this month. When a company wants to keep the identity of a gene secret, itwill give it a code name and a general description, such as "Gene S is ahormone in humans. It is harmless to invertebrates," and so on.

He also says the only drugs so far being grown in crops are proteins that wouldsimply be digested if accidentally eaten by humans or animals. "The risksare minimal," White says. "No one is making Viagra in a field."But there's nothing to stop companies producing a Viagra crop if they want to."It's a disaster waiting to happen," says Doreen Stabinsky, a scienceadviser for Greenpeace. "Grow this stuff in a greenhouse or a cave, not inan open field where animals can grab the seeds."

Ellstrand agrees that stricter containment is needed. Pharmed plants could begenetically engineered to prevent gene flow using methods such as the infamousTerminator technique, which makes seeds sterile, or a newly proposed one dubbedthe Exorcist (see "Begone! evil genes"). And to be absolutely certainthe food supply is safe, he argues that only plants that aren't grown for foodshould be used to make drugs.

White points out that when the long-awaited regulations are finally published,the public will have 120 days to respond. "I wouldn't be surprised if wegot thousands of letters telling us: not in food."