What happened?

On Friday September 22, Kraft Foods announced a voluntary recall of Taco Bell brand taco shells because the tacos likely contain a protein from a variety of insect-resistant corn that is approved for animal feed but not yet for human food.

Why was the food recalled?

The recall was made by Kraft because the tacos "contained a product that is not lawfully to be in food, but not because it is a major health risk," stated Joseph Levitt of the FDA at a hearing September 26 to the US Senate Committee on Health, Education, Labor and Pensions. In this case, for several months the EPA has been reviewing data on the protein which, while not considered toxic, is still being assessed for its potential to trigger an allergic response.

If the food is not a major health risk, then why did Kraft recall it?

In general, food producers may have several reasons to recall a product. A product may be recalled because it contains a product that is not yet approved for use even though it is not an established threat to health. That is the case with the Taco Bell tacos. A product may be recalled if it contains an ingredient that is safe and approved but is not intended to be in that product, or if the product contains an unsafe component. Food producers may also recall a product to sustain public confidence and to protect their brand name.

Who detected the protein? 

The detection of the protein was announced September 18 by the Genetically Engineered Food Alert coalition and reported in that day's edition of the Washington Post. Later in the week results of tests run on behalf of Kraft led the company to issue the recall.

What is the protein in corn that is approved for feed but not for food? 

The protein, Bt Cry9C, makes corn resistant to certain insect pests. It is present in one variety of genetically-engineered corn called StarLink by the biotech company Aventis.

What is Bt protein? 

Bt Cry9C protein is one of many members of the "Bt protein" family used by organic growers and in biotech crops to control pests. Bt proteins come from various strains of a bacterium called Bacillus thuringiensis. Bt can be applied as a powder of dried bacteria, or the gene for a Bt protein can be added to the genes of a crop so that a plant makes its own Bt protein. At least eight varieties of corn that make other versions of Bt protein have been reviewed and approved by the EPA and are in the American food supply. These use versions of Bt protein that have been present before in the food supply from their use as a pesticide applied as a powder.

Why is the allergenicity of Cry9C in question? 

The Cry9C version of Bt has not been in the food supply before, and like any protein new to the food supply, no one knows for sure how the immune systems of people will respond to it. Part of the governmental review process includes an assessment of a new protein's potential to cause allergies. The concern about allergenicity is based on the fact that, unlike most other Bt proteins, the Cry9C protein is resistant to degradation by heat, acid, and proteases (enzymes that chew up many proteins). The Washington post article reported that a panel of scientific advisors to the EPA concluded in July that "there is no evidence to indicate that Cry9C is or is not a potential food allergen.

Why is the corn approved for animal feed but not for human food? 

The EPA concluded that the protein was effective in fighting pests and sufficiently safe for the environment, but the allergenicity question was open and still being studied further. Therefore, EPA approved the corn for use as animal feed but not for human food, and Aventis established a program to work with farmers to keep the corn variety limited to use animal feed.

Are such approvals for feed but not food common?

As far as I know, it's the only case. And since it created a situation that invites unintended mixing of corn that's approved and with corn that is unapproved for food, several organizations, from Kraft to the Biotechnology Industry Organization, have called on the EPA to stop the practice of approving crops for feed but not for food. And such split approvals present an Achilles heel: while farmers may try to keep corn approved only for feed separate from other corn approved for human food, some mixing is likely to happen.

Furthermore, biotechnology techniques such as DNA fingerprinting by PCR (polymerase chain reaction) enable people to find even tiny levels of mixing. Since the EPA has not established any levels of tolerance for such mixing, corn approved only for feed is liable to ruin the value of other corn for human use. One way to reduce that possibility is to make it a policy that all corn grown for feed must also be approved for food.

On September 27, Aventis announced that it was suspending seed sales of the StarLink variety of corn and will not resume seed sales until the corn becomes approved by the EPA for human food.

Isn't there a system in place to prevent mixing of unapproved corn with approved corn?

It is one thing to have regulations and systems in place and intended to prevent mixing of approved and unapproved corn. It is another thing to assume that the preventions will be perfect. A homeowner is smart to take steps to prevent leaks in a home's plumbing, and even smarter to anticipate that leaks may happen despite the preventions. The legal remedy when an unapproved protein leaks into the food supply is to issue a recall.

Why can't regulators decide if the protein is an allergen? 

Because scientists do not yet know reliably what makes some proteins trigger allergies, and other proteins not trigger allergies. Foods that commonly cause allergies are also common foods: milk, beans including soybeans and peanuts, tree nuts, rice and shellfish are examples. The allergic response depends both on the triggering allergen and the immune response of an individual. Why some people get allergic to some foods while others don't is also not well understood.

Furthermore, the absence of evidence of allergenicity does not necessarily mean that a protein will never cause an allergic reaction in some people. Because of all the variables involved, projecting whether a compound will be an allergen is not yet an exact science; it's more like predicting the weather in a year than predicting a lunar eclipse.

If gene-spliced crops are avoided, will we reduce the risk of new allergens? 

Maybe, and maybe not. Plant breeders develop new crop varieties every year using conventional breeding. The new varieties are selected to resist pests or diseases or drought or yield more or grow better. At the molecular level, it is likely that sometimes new proteins never before in the food supply are responsible for making the new traits. The difference is that in conventional breeding the scientist usually has no idea what the molecular basis is for the new trait, while with biotechnology-bred crops, scientists can study at least the one new protein added to the crop. So while there's likely to be a new protein or proteins in at least some of the new crop varieties with new traits developed through conventional breeding, these new proteins will go untested not because they are allergy-free but because they are unknown. Only biotech varieties will be tested because only biotech varieties have a known target protein to suspect. This can lead to the mis-impression that the risk is unique to biotech varieties, when in fact the risk of new allergens is present in all new varieties. While the risk is widespread, among all new crops only biotech crops get the scrutiny.

Does the system for regulating foods derived from crops improved by biotechnology need tuning, fixing or overhauling? 

That question is under debate. Some people propose special labeling as an added safeguard. Other people argue that it is misleading to continue to add special Īsafeguards' targeting biotech crops when the National Academy of Science has repeatedly concluded that biotech crops pose no special risks compared to conventional crops. The FDA stated clearly on May 3 that it is making its currently voluntary review system mandatory­not because of extra risks, but rather to generate public confidence. This leads to a paradox: if most scientists say that biotech crops are as safe as conventional crops, then why does the government regulate them differently? Clearly one reason is to assuage the public. The irony is that additional regulations designed not to reduce risks but to reassure the public might end up doing neither.