Bt toxins are superb for pest control because they are lethal to certain insects, yet harmless to most other organisms, including people. They bind to specific target sites in insect midgut membranes and kill by disrupting the membranes. Because of their specificity, Bt toxins are environmentally friendly alternatives to conventional insecticides. For many decades, organic growers and others have sprayed commercial formulations of Bt toxins for pest control.

  What about pest resistance to Bt toxins?

Despite their widespread use in sprays for decades, no cases of resistance to Bt toxins were reported in field populations of insect pests before 1990. This contrasted sharply with the thousands of documented examples of resistance to other insecticides in more than 500 species of insects. The picture changed when we discovered that some diamondback moth populations in Hawaii evolved resistance after repeated treatments with sprays of Bt toxins. Since then, many Bt-resistant field populations of diamondback moth have been found in Asia and North America. Laboratory selection has produced Bt-resistant strains of more than a dozen other pests, demonstrating that the genetic potential to evolve resistance to Bt is not limited to diamondback moth.

  Does production of Bt toxins in transgenic crops make pest resistance more likely?

Yes. In the late 1980s, crop plants that produce Bt toxins were created by incorporating the genes encoding Bt toxins into the plants' genomes. In effect, this enabled the plants to produce their own environmentally friendly insecticide and reduced the need for sprays. Three Bt crops were commercialized in the U.S.—Bt corn, Bt cotton, and Bt potato. Widespread planting of Bt crops makes pest resistance more likely because it increases pest exposure to Bt toxins in time and space. Unlike the Bt toxins in sprays—which are degraded by sunlight in hours or days—the Bt toxins in transgenic crops are produced throughout the season. Bt crops planted on millions of hectares also vastly increase the area over which pests are exposed to Bt toxins.

  How do Bt crops change the risk-benefit analysis of Bt as an environmentally friendly insecticide?

Because Bt toxins in sprays degrade so quickly, they can be relatively expensive and ineffective compared with conventional insecticides and other control tactics. On the other hand, the rapid decay reduces selection for pest resistance and limits accumulation in the environment and potential side effects. Production of Bt toxins by transgenic crops makes the economics much more favorable because control may be effective for months. Also, Bt corn and Bt cotton kill pests that bore inside plants where they are shielded from sprays. But this approach selects for resistance all season long. It also raises other issues associated with the genetic engineering of crop plants, including impacts of Bt toxins on soil organisms and movement of Bt genes from Bt crops to weeds.

  What is it about your 1997 PNAS paper that gave it such impact?

The paper reported results affecting perceptions about how long Bt crops might be effective. The controversy about transgenic crops raised the stakes about the possibility of pest resistance. If pests were to quickly evolve resistance, this would limit the usefulness of Bt crops. Thus, critics of biotechnology seized on resistance as a major drawback of Bt crops. Meanwhile, to inform the debate with scientific evidence, research on Bt resistance accelerated worldwide.

  What was the major finding reported in the paper?

We found that one resistance gene in diamondback moth conferred resistance to four Bt toxins. This means that a population exposed to only one toxin would evolve resistance to that toxin and cross-resistance to three others. The four toxins we studied are closely related. Resistance did not extend to distantly related Bt toxins. So, selection with a single Bt toxin can produce cross-resistance to a cluster of closely related Bt toxins, whereas cross-resistance to less similar Bt toxins is less likely.

  What is the significance of this finding?

In the Bt crops grown commercially so far, each individual plant produces only one type of Bt toxin. In principle, if cross-resistance does not occur, resistance can be delayed by simultaneous or sequential deployment of more than one Bt toxin. The type of cross-resistance reported in the paper reduces the number of options available for countering pest resistance.

  What was the biggest challenge to doing this study?

Although it is a standard technique in quantitative genetics, testing families from single pairs of parents was not done much in resistance studies at the time. To generate each family, we paired a single female with a single male and allowed them to mate. We reared their offspring as a family. We split each family and tested each subset of offspring from a family with a different toxin. If a family was resistant to one toxin, it was almost always resistant to all four. This was a new and compelling way to demonstrate cross-resistance.

  Was there anything particularly surprising about your results?

By testing offspring from single-pair crosses between a susceptible strain and a resistant strain, we estimated that 21% of individuals from the susceptible strain were heterozygous for the gene that conferred resistance to four Bt toxins. This means that the resistance gene was at about 10%, which is 100 times higher than expected. Put this together with the finding that a single gene can confer resistance to four toxins and it suggests that resistance could evolve rapidly in this species. Indeed, in many places, this pest has evolved resistance to sprays of commercial products containing three of the toxins we studied.

  Was the paper cited so heavily because the results were so surprising or because the topic itself was so hot?

I guess it was both. Our results gave more credence to the idea that resistance could evolve in insects targeted by genetically engineered crops. Commercially grown Bt crops do not target diamondback moth. But given that diamondback moth can evolve resistance to Bt toxins in the field and pests targeted by Bt crops can evolve resistance in the lab, how long will it be before resistance to Bt crops occurs in the field?

  Where has this research gone since 1997?

In collaboration with other research teams, the work has gone in several directions. We found that the resistance extended beyond the initially studied four toxins to at least five toxins (Cry1Aa, Cry1Ab, Cry1Ac, Cry1Fa, and Cry1Ja). We mapped the chromosomal location of the Bt resistance gene in the strain from Hawaii. We also learned that a strain of diamondback moth from Pennsylvania had a resistance mutation that was virtually identical to the one from Hawaii. We are now conducting parallel research on pink bollworm, a major pest targeted by Bt cotton. A primary goal is to develop DNA testing to monitor resistance in field populations. This could help to improve understanding and management of resistance.

  Is there a broader message from your research about the controversy over genetically engineered crops?

Bt crops were first grown on a large scale in the United States in 1996. Based partly on the experience with diamondback moth, some scientists predicted that resistance to Bt crops would occur in a few years. This hasn't happened. Many field populations of diamondback moth have evolved resistance to Bt sprays, but resistance in the field to Bt crops has not yet been detected. We don't know if resistance to Bt crops will occur in 2002 or not for another six years or more. We do know that the worst-case scenarios have already proved too pessimistic.

Bruce Tabashnik, Ph.D.
Department of Entomology
College of Agriculture and Life Sciences
University of Arizona
Tucson, AZ, USA