Safe Food: The Politics of Food Safety (32 page)

One especially ironic aspect of this situation is that the food biotechnology industry, in achieving an unregulated marketplace, made itself vulnerable to charges of producing allergens (through lack of testing) and covering up the hazards of transgenic foods (through lack of labeling). The StarLink corn episode illustrates this irony. Pioneer Hi-Bred knew in 1996 that soybeans intended for chicken feed could not possibly be kept separate from those intended for human consumption. As described in the introductory chapter, both Aventis and the Environmental Protection Agency (EPA) ignored this lesson at great cost. With their success in finding evidence of StarLink corn in common food products and
revealing gaps in the regulatory system for genetically modified foods, advocates could use allergenicity—a safety issue—as a means to oppose the industry’s economic and political goals. StarLink’s owner could not demonstrate the safety of the corn to the satisfaction of EPA advisory committees and was forced to withdraw it from the market, albeit too late.
²⁰
Supposedly scientific arguments about the degree to which transgenic foods might be allergenic reflect underlying concerns—less easily debated—about who is entitled to decide what people eat.

A second legitimate safety issue is antibiotic resistance. In
chapter 1
, we saw how the routine use of antibiotics as growth promoters in cows and chickens favored the emergence of resistant microbial pathogens, rendering the antibiotics useless against human infections. Plant biotechnology raises similar concerns. In creating new plant varieties, agricultural biotechnologists link genes for antibiotic resistance to the genes they want to transfer into plants; these genes act as selection markers to identify the rare plants that actually accept the new genes. This selection system works because the plants that take up the antibiotic-resistance marker genes are the only ones to survive when grown in a broth containing the antibiotics (see appendix). Given the importance of antibiotic resistance as a public health problem, it makes sense to ask whether genetically engineered foods contribute to that problem. Answering that question requires a brief discussion of how antibiotics work.

Molds and bacteria naturally produce chemicals—antibiotics—that interfere with the growth or reproduction of
other
bacteria but are not nearly so toxic to animals or humans. Antibiotics act by blocking specific steps in the synthesis of structures or in metabolic processes unique to bacteria: cell walls (penicillin), cell membranes (polymyxin B), proteins (streptomycin, chloramphenicol, tetracycline), nucleic acids (rifampin), or folic acid vitamins (sulfonamide, trimethoprim). When animals or humans take antibiotics appropriately—in the right dose for the right length of time—the drugs suppress the growth of all sensitive bacteria. Bacteria, however, are exceedingly small, and the normal digestive tract contains hundreds of billions of them. Among this multitude, some are likely to lack the target structure; these grow in the presence of the antibiotic. Penicillin, for example, has no effect on bacteria that lack cell walls. Bacteria can acquire antibiotic resistance by mutations that change the structure of DNA and favor survival or produce enzymes that destroy
the antibiotics or pump them out. The use of low-dose antibiotics “selects” for such bacteria; the drugs kill off most competing bacteria and allow the resistant ones to proliferate.
²¹

The use of marker genes for antibiotic resistance in plant biotechnology raises additional concerns. Perhaps the genes for such characteristics will jump to other bacteria, and the bacteria will become resistant to
multiple
antibiotics. Scientists transfer new genes into plants by using special pieces of bacterial DNA called plasmids. Plasmids often contain three kinds of genes relevant to this discussion: (1) genes that enable them to “infect” and transfer selected genes into plants, (2) genes for antibiotic resistance, and (3) genes that enable them to infect many different kinds of bacteria (see appendix). Plasmid-containing bacteria in the intestines of animals or people could transmit antibiotic resistance to other bacteria, some of which might be pathogenic. This possibility is not just theoretical. Some pathogenic bacteria once easily controlled by penicillin are now thoroughly resistant to that drug, and others in ground meat have been found to resist treatment by as many as 12 antibiotics.
²²

Such findings explain why the continued use of low-dose antibiotics in farm animals elicits so much concern. They also explain why health officials want food biotechnologists to stop using clinically important antibiotics as selection markers. They want to avoid any chance—no matter how improbable—that transgenic plants might “lose” their recombinant antibiotic-resistance markers and transmit them to soil bacteria, to animals, or to people. In the worst-case scenario, a plant gene might recombine with the DNA of bacteria living in the intestines of animals or people and pass the trait for antibiotic resistance along to disease-causing bacteria. The antibiotic used in the selection process would then be ineffective as a treatment option. Alternatively, the antibiotic might be useless if people taking it were eating foods containing genes for resistance to that drug.

Perhaps because most scientists believe that such possibilities are exceedingly remote, the question of antibiotic-resistance markers also exists in a regulatory vacuum. Attempts to regulate transgenic antibiotic resistance began in 1990, when Calgene, an agricultural biotechnology company, asked the FDA for an opinion about whether it could use a gene for resistance to the antibiotic kanamycin (neomycin) as a selection marker for constructing transgenic tomatoes and canola oilseeds. This particular resistance gene specifies production of an enzyme able to inactivate kanamycin and related antibiotics. By the time the FDA issued its 1992 policy on genetically engineered plants, kanamycin
already
was
in use as a selectable marker for development of more than 30 transgenic crops. In that policy, the FDA made no particular recommendation about antibiotic-resistant marker genes but said that its scientists were evaluating the issue.
¹⁸
In 1993, hoping to elicit a decisive response, Calgene petitioned the FDA to permit use of the kanamycin-inactivating enzyme as a “food additive” in genetically modified foods and cotton. In 1994, the FDA convened a meeting of its Food Advisory Committee to consider the Calgene petition.

I was one of four consumer representatives to that committee at the time, all of us united in what turned out to be the minority opinion. We were troubled by the lack of satisfactory answers to our questions about the probability of transferring antibiotic resistance. We urged caution but were heavily outvoted. After the meeting, FDA officials correctly reported that committee members “generally” approved the agency’s regulatory approach and agreed that Calgene had addressed the relevant scientific questions. Thus, the FDA ruled that Calgene’s evidence met the legal definition of safety for food additives: reasonable certainty that no harm would result from use.

Calgene contended that the kanamycin-inactivating enzyme (like all enzymes, a protein) would be destroyed by cooking or normal digestive processes and was unlikely to function in the intestine. But could the
gene
for antibiotic resistance jump from food or soil to bacteria in the intestines of animals or people? The FDA considered this suggestion too highly improbable to be worth discussion. In approving the kanamycin-inactivating enzyme as a food additive, the FDA explained that its policy is not to regulate genes or DNA: “DNA is present in the cells of all living organisms, including every plant and animal used for food by humans or animals, and is efficiently digested. . . . The DNA that makes up the [kanamycin-resistance] gene does not differ from any other DNA and does not itself pose a safety concern.”
²³

In its decision, the FDA emphasized that safety “does not—and cannot—require proof beyond any possible doubt that no harm will result under any conceivable circumstance.” Nevertheless, the agency agreed to consider further requests for use of selection markers for resistance to other antibiotics on a case-by-case basis. Subsequently, various groups challenged the FDA about the safety and regulatory status of antibiotic-resistance marker genes and, in late 1996 and early 1997, the agency consulted with outside experts about whether the use of such genes might cause problems. On the basis of those discussions, the FDA drafted a guidance statement for industry. This reassuring document said that antibiotic-resistance
genes in food were “not of great concern,” as the chance that they might be transferred from plants to bacteria in the intestine or environment was “remote.”
²⁴

It is difficult to know how to interpret the FDA’s decisions or guidance suggestions. Either transgenic transfer of antibiotic resistance is a problem or it is not. The FDA’s use of the word “remote” suggests that marker genes require no special attention, but its nonbinding guidance document advises developers to evaluate the use of these genes quite carefully. Developers of new plant varieties, according to the FDA, should find out whether their marker genes involve clinically important antibiotics and whether they could transfer resistance to bacteria. If so, developers should use something else. Furthermore, if an antibiotic is the
only
one available to treat a particular infection in animals or people, it should not be used at all.

The inherent ambiguity of the agency’s position seemed certain to—and did—elicit contentious comments, but the FDA did not respond to them. In the meantime, countries throughout Europe used concerns about antibiotic resistance as a basis for bans on the development and growth of transgenic food plants, and U.S. groups also used the issue to raise objections about food biotechnology. In 2001, when the Department of Health and Human Services (DHHS), the FDA’s parent agency, released the
Action Plan to Combat Antimicrobial Resistance
, its principal recommendation to prevent such problems was a public education campaign to reduce the clinical use of antibiotics—not to reduce antibiotic use in animals.
²⁵

Overall, the relaxed regulatory environment demanded by the food biotechnology industry raises many of the outrage issues listed in
table 2
. No matter how remote the health hazards might be, the industry’s antiregulatory stance does little to inspire trust. If anything, the stance invites criticism on safety and other grounds. As we will now see, similar considerations affect issues related to the environmental effects—risks and benefits—of genetically modified foods.

The mandate of the FDA is to assure the safety of drugs, medical devices, and foods, and the agency’s policy for food biotechnology focuses on consequences that might present direct risks to
human
health. In approving transgenic foods, the FDA does not consider whether they might pose
ecological
risks. They might, for example, displace existing plants and
animals, create new plant pathogens, disrupt ecosystems, transfer genes to weeds or wild relatives, reduce crop diversity, or “contaminate” native plants or organically grown foods. Widespread planting of
Bt
crops, for example, might encourage the proliferation of insects resistant to the
Bt
toxin. Similarly, widespread use of herbicide-resistant crops might transfer that resistance to undesirable weeds or encourage further reliance on chemicals—such as Monsanto’s Roundup—as pest-management strategies.
²⁶
Despite such concerns, plantings of transgenic crops increased from negligible acreage in 1995 to hundreds of millions of acres within just a few years. Agricultural producers quickly adopted transgenic soybeans, corn, and cotton, largely because they simplify the control of weeds and insect pests by requiring fewer applications of the more toxic chemicals. Farmers, apparently, perceive significant benefits from growing transgenic crops, but how are we, as citizens and consumers, to reconcile the risks and benefits? Let’s begin by looking at the risks.

When researchers began to examine questions of environmental risk, their early results provided plenty of justification—albeit highly preliminary—for concern. In 1996, for example, farmers planted 2 million acres with Monsanto’s
Bt
cotton, but lost thousands of acres when the toxin failed to protect against a bollworm infestation. This event raised the uncomfortable possibility that such huge plantings might promote
Bt
resistance.
²⁷
According to investigative accounts, EPA officials asked Monsanto to evaluate whether the surviving bollworms were indeed
Bt
-resistant, but the agency could not force the company to cooperate: “Further evaluation of the crop is entirely dependent on Monsanto’s own reporting.”
²⁸
Also in 1996, researchers reported that transgenic oilseed (canola) plants readily transmitted herbicide resistance to related weeds. Because weeds reproduce rapidly and compete for nutrients with crop plants, this finding raised fears that cross-pollination might create herbicide-resistant “superweeds” that could overrun cropland and cause an ecological catastrophe. EPA officials revealed the consequences of the regulatory gap, however, when they explained that monitoring of herbicide resistance is not a federal responsibility: “It is the developer of the product that has the interest in assuring that resistance does not build up.”
²⁹

These early reports on environmental risks were based on single studies and needed further confirmation, but others soon followed. For example, preliminary studies showed that bees and other beneficial insects
die when exposed to the
Bt
toxin, but certain harmful moths and tobacco budworms resist it. The
Bt
toxin remains stable in soil for many months, meaning that it exerts continuous pressure to encourage the growth of resistant insects. Herbicide-resistant plants transfer resistance to related weeds, sometimes over great distances through pollen drift.
³⁰
Many such problems can be unintended consequences of large-scale plantings of transgenic crops, and they greatly trouble environmentalists. As I will soon explain, effects on monarch butterflies are the most political of such consequences, but let’s look first at the environmental
benefits
claimed for transgenic crops.