A key issue with genetically modified crops is how to prevent engineered traits, such as pesticide resistance, from spreading to other crops or weeds. Now a proof-of-principle trial suggests a simple solution: GM pollen could contain an additional gene that would kill any weedy offspring. The GM crops, however, would reproduce happily amongst themselves, allowing farmers to save and replant seed.
Researchers at the government agency Agriculture and Agri-Food Canada, in Ottawa, tested the idea by outfitting tobacco plants with two genes from Agrobacterium tumefaciens, a microbe that causes crown gall disease. These genes unleash a surplus production of the hormone auxin, which wreaks havoc with an infected plant. The team made sure the damage would only afflict seeds by linking the genes to a genetic switch, active only in embryos, from another kind of plant. When the researchers put this genetic assembly into tobacco plants, it killed their seeds.
To keep this weapon in check and allow the GM plants to produce viable seeds, the team added another genetic control switch that repressed the auxin genes. The safety catch is released when the plants breed with nonengineered relatives, however. Because of how the genes were added to the chromosomes, the GM pollen only contains the repressor or the seed-killing genes--but not both. When the GM tobacco plants were bred with nonengineered relatives, all of the offspring that inherited just the seed-killing genes failed to germinate, the team reports online this week in the Proceedings of the National Academy of Sciences. "It's a proof of concept," says team-member Johann Schernthaner, a molecular geneticist. He says that the technique would be simpler to use than some of the other gene-containment methods, such as GM plants that must be sprayed with chemicals before they can produce seed.
"This looks like a real breakthrough, assuming certain technical challenges can be overcome," says gene-flow expert Allison Snow of Ohio State University, Columbus. For instance, as Henry Daniell of the University of South Florida in Tampa notes, the repressor itself would need to be modified so that it's lethal when appearing alone in pollen. Given the technical difficulties of constructing such systems, Daniell says that he's a bigger fan of creating GM crops with new traits engineered into their chloroplasts, which in most major crops don't spread via pollen.