Insects Develop Resistance to Engineered Crops
when single- and double-gene altered plants are in proximity, Cornell researchers say
By Krishna Ramanujan
June 17, 2005
ITHACA, N.Y. -- Genetically modified crops containing two insecticidal proteins in a single plant efficiently kill insects. But when crops engineered with just one of those toxins grow nearby, insects more rapidly develop resistance to all the insect-killing plants, report Cornell University researchers.
A soil bacterium called Bacillus thuringiensis (Bt), whose genes are inserted into crop plants, such as maize and cotton, creates these toxins that are deadly to insects but harmless to humans.
Bt crops were first commercialized in 1996, and scientists, critics and others have been concerned that widespread use of Bt crops would create conditions for insects to evolve and develop resistance to the toxins.
Until now, it has not been shown if neighboring plants producing a single Bt toxic protein might play a role in insect resistance to transgenic crops expressing two insecticidal proteins.
"Our findings suggest that concurrent use of single- and dual-gene Bt plants can put the dual-gene plants at risk if single-gene plants are deployed in the same area simultaneously," said Anthony Shelton, professor of entomology at Cornell's College of Agriculture and Life Sciences and an author of the study, which was posted online June 6 in the Proceedings of the National Academy of Sciences (PNAS) and is in the June 14 print edition of the journal. "Single-gene plants really function as a steppingstone in resistance of two-gene plants if the single gene plants contain one of the same Bt proteins as in the two-gene plant."
Cotton and maize are the only commercial crops engineered with Bt genes. In 2004 these crops were grown on more than 13 million hectares (about 32 million acres) domestically and 22.4 million hectares (more than 55 million acres) worldwide. After eight years of extensive use, there have been no reports of crop failure or insect resistance in the field to genetically modified Bt crops, Shelton said. Still, several insects have developed resistance to Bt toxins in the lab, and recently, cabbage loopers (a moth whose larvae feed on plants in the cabbage family) have shown resistance to Bt sprays in commercial greenhouses.
So far, only diamondback moths, which were used in this study, have developed resistance to Bt toxins in the field. The resistance resulted from farmers and gardeners spraying Bt toxin on plants for insect control, a long-standing practice. While Bt toxin sprayed on leaves quickly degrades in sunlight and often does not reach the insect, genetically modified (GM) Bt plants express the bacterium in the plant tissue, which makes Bt plants especially effective against insects that bore into stems, such as the European corn borer, which causes more than $1 billion in damage annually in the United States.
In greenhouses at the New York State Agriculture Experiment Station in Geneva, N.Y., the researchers used three types of GM broccoli plants: two types of plants each expressed a different Bt toxin, and a third -- known as a pyramided plant -- expressed both toxins. Elizabeth Earle and Jun Cao, co-authors of the PNAS paper and members of the Department of Plant Breeding and Genetics at Cornell created the plants.
For their studies, the researchers employed strains of diamondback moth that were resistant to each of the Bt proteins. The combination of Bt plants and Bt-resistant insects allowed them to explore the concurrent use of single- and dual-gene Bt plants in a way that could not be done with cotton or maize, although their results are relevant to these widely grown plants.
First, the researchers bred moth populations in which a low percent of the moths were resistant to a single Bt toxin. The insects were then released into caged growing areas with either single-gene plants, dual-gene plants or mixed populations and allowed to reproduce for two years.
The researchers found that after 26 generations of the insect living in the greenhouse with single-gene and dual-gene plants housed together, all the plants were eventually damaged by the insects, because over time, greater numbers of insects developed resistance to the plants' toxins. However, in the same two-year time frame, all or almost all of the insects died when exposed to pyramided plants alone.
"It's easier for an insect to develop resistance to a single toxin," said Shelton. "If an insect gets a jump on one toxin, then it becomes more rapidly resistant to that same toxin in a dual-gene plant. And when one line of defense starts to fail, it puts more pressure on the second toxin in a pyramided plant to control the insect," Shelton added.
While single-gene Bt plants are most prevalent, industry trends suggest that pyramided plants may be favored in the future. In Australia, the use of single-gene Bt cotton was discontinued two years after farmers began planting dual-gene cotton in 2002. In the United States, companies introduced dual-gene cotton in 2003, but single-gene varieties remain on the market.
"Single-gene Bt plants have provided good economic and environmental benefits, but from a resistance management standpoint they are inferior to dual-gene plants. U.S. regulatory agencies should consider discontinuing the use of those single-gene plants as soon as dual-gene plants become available," Shelton said. "And industries should be encouraged to create more dual-gene plants."
Along with effective insect control, pyramided plants have an added advantage of requiring a smaller refuge -- a part of the field where non-Bt plants are grown. Refuges create opportunities for Bt-resistant insects to mate with other insects that do not have resistance. The offspring of such a mating will be susceptible to the toxins.
"Having a refuge is a good management strategy, but it is not suitable for small farmers in China and India," said lead author Jian-Zhou Zhao, a senior research associate in entomology at Cornell. "The two-gene strategy is more suitable in developing countries like China where farmers have an average of half a hectare (1.2 acres) of land, much less land than American farmers, and not enough to spare for refuges."
A U.S. Department of Agriculture Biotechnology Risk Assessment Program grant supported the study.