Update on Risk Research

Process Counts
For years, biotech companies have maintained that the actual process of genetic engineering doesn't affect safety. Only spliced-in traits matter. For example, tomatoes made herbicide-resistant by two different processes--genetic engineering and traditional methods--would be equivalent products and equally safe to eat.

Joy Bergelson and colleagues at the University of Chicago recently reported experiments in Nature that challenge that assumption. Their work with wild mustard plants suggests that genetic engineering itself, apart from the new genes added, can cause dramatic changes in a transgenic plant.

The scientists obtained two versions of herbicide-resistant mustard plants--both containing the same gene for herbicide resistance--but derived by two methods--genetic engineering and traditional breeding. They then compared rates of gene flow to nearby relatives from the two sets of resistant plants. What they found was stunning--genetically engineered herbicide-resistant plants were 20 times more likely to outcross, that is, interbreed with relatives, than traditionally produced ones.

How the engineering process operates to alter outcrossing rates remains a mystery.

Source: J. Bergelson et al., "Promiscuity in transgenic plants," Nature 395: 25, September 3, 1998.

More on Transgenes in Wild Populations
Another study reported this summer challenges another line of argument used by industry in the safety debate: that novel genes transmitted to wild populations of plants would likely die out because the plants would not be able to compete with wild counterparts. Allison Snow, an Ohio State University ecologist, reported that wild plants containing a transgene for herbicide resistance can hold their own in competition with their unmodified counterparts.

Snow's research demonstrates that wild mustards containing the herbicide-resistance gene produce just as many seeds as sibling mustards without the transgene, even when no herbicide is applied to the plants.

Conventional wisdom has held that an herbicide-resistance transgene will not stay in a wild population, and, in fact, would be detrimental to survival, unless the plants are sprayed with the herbicide. An herbicide-resistance gene was expected to decrease survival except where the herbicide was present, perhaps because of the energy cost in maintaining an unneeded gene. Lowered survival would eventually lead to the loss of the transgene from the population.

Sources: A. Snow and R. Jorgensen, "Costs of transgenic glufosinate resistance introgressed from Brassica napus into weedy Brassica rapa," Abstract of a paper presented at the annual meeting of the Ecological Society of America, Baltimore, Md., August 6, 1998; A. Dove et al., "Research news: promiscuous pollination," Nature Biotechnology 16: 805, September 1998.

Buildup of Active Bt Toxins in Soil
Research from New York University indicates that active Bt toxins genetically engineered into crops may accumulate in soil. In laboratory experiments, Guenther Stotzky and his colleagues have shown that purified Bt toxins, similar to ones found in some lines of transgenic Bt crops, do not disappear when added to soil but instead become rapidly bound to clay and humic acid soil particles. The bound Bt toxins, unlike free toxins, are not degraded by soil microbes, nor do they lose their capacity to kill insects.

The accumulation of active Bt toxins in soils could represent a risk to soil ecosystems. Typically, toxins in naturally occurring Bt bacteria, and sprays made from them, are not active--they exist in the form of inactive, so-called protoxins. Before they are able to kill an insect, the protoxins must be dissolved in its gut and cut by protein-digesting enzymes liberating the active toxins. By contrast, the toxin in many Bt crops needs no activation. It is already in an active form.

Stotzky suggests that active Bt toxins might be released to the soil as farmers incorporate plant material into the ground after harvest. Active toxins, which might build up with time, could kill known Bt-sensitive soil insects. In addition, a broader range of insects and other organisms may be susceptible to engineered toxins than to toxins from naturally occurring bacteria. Organisms unable to dissolve or cut the protoxin but sensitive to the active toxin, would not be harmed by the bacterial toxin but would be vulnerable to the engineered active form. Soil-inhabiting insect pests, already exposed to the toxin in their plant-eating phase, may be under continuing pressure to evolve resistance to Bt.

Stotzky's results, if they hold true under field conditions, should sound an alarm to regulators and others concerned about the risks of genetically engineered crops. To the extent that Bt crops containing active toxins are planted in the United States, soil organisms may be newly exposed to active Bt toxin. Sprays contribute far less active toxin to soil ecosystems because, for the most part, they exist in an inactive form. In addition, unlike the engineered toxins, which are protected inside the plant, spray toxins on the surfaces of leaves and soil are subject to inactivation by UV light before they have a chance to be incorporated into soils.

Sources: C. Crecchio and G. Stotzky, "Insecticidal activity and biodegradation of the toxin from Bacillus thuringiensis subsp. kurstaki bound to humic acids from soil," Soil Biology and Biochemistry 30: 463-70, 1998, and references therein.

Intoxicating Bacterium Kills Plants
A recent report in Applied Soil Ecology illustrates the unexpected ways in which environmental release of genetically engineered microorganisms might cause widespread ecological damage. The core experimental finding is that the addition of a genetically engineered bacterium, Klebsiella planticola (SDF 20), to a small microcosm consisting of wheat plants and sandy soils kills the plants, while the addition of the non-engineered parent, Klebsiella planticola (SDF 15), does not.

Klebsiella (SDF 20) is a lactose-fermenting bacterium engineered to produce increased ethanol concentrations in fermentors that convert agricultural wastes to ethanol. The system (developed in Germany) envisioned the disposal of fermentation residues, including the engineered bacteria, as an organic soil amendment. The report that the engineered bacteria cause plant death raised the possibility that soil amendments would kill or impair crops in the fields where they were used and, further, that, once released and established, the Klebsiella could not be eliminated.

The paper explored but failed to nail down the mechanism of plant killing. Whatever the mechanism, the research suggests that engineered microorganisms can have far reaching, potentially devastating, effects.

Source: M.T. Holmes et al., Effects of Klebsiella planticola on soil biota and wheat growth in sandy soil. Applied Soil Ecology 326: 1-12, 1998.

Go to UCS USA Gene Exchange contents

Union of Concerned Scientists
2 Brattle Square, Cambridge, MA 02238-9105
617-547-5552, ucs@ucsusa.org