As discussed in Chapters 3 and 10, high-dose acute toxicology studies are required by the U.S. Environmental Protection Agency (EPA) to assess the potential hazards of plant-incorporated protectants (PIPs). This testing requirement is based on the need to demonstrate that the toxic mechanism of the plant protectant is not relevant to animals and man. For example, the knowledge that existing commercial insecti-cidal Cry proteins (derived from Bacillus thuringiensis bacteria) act through acute mechanisms at low doses to control insect pests (Chapter 3) and that does not occur in man is important and reassuring from the safety perspective. The EPA requires that PIPs be tested at high dosage levels (generally g/kg body weight where feasible) to confirm their safety. Further, although most consumed proteins are not toxic, those that are toxic generally exert their effects through acute modes of action.82
The procedures for carrying out high-dose acute testing of proteins were presented in Chapter 10. To date, no treatment-related adverse effects have been observed up to the highest dosages tested (Table 11.1). As will be shown later, the high dosages of proteins administered to mice are orders of magnitude higher than potential human dietary exposures from consuming food from biotechnology-derived crops. For PIPs that have a history of safe use and defined mode of action, the EPA does not require additional toxicology testing beyond acute oral maximum hazard dose testing.22
Acute toxicology studies are generally conducted via the oral route because the diet is the most likely route of human exposure to the proteins introduced into biotechnology-derived crops. Mice are generally used instead of rats as they are approximately 1/10 the body weight of rats and require much less protein for dosing. Mice are also known to be sensitive to the adverse effects of known protein toxins and are most commonly used to assess their toxic effects.83
Intravenous (IV) dosing has also been used to assess the intrinsic safety of proteins introduced into biotechnology-derived crops.41 Generally, low dosages (~10 mg/kg) of the introduced protein are administered as it is assumed that only small amounts of ingested proteins could be absorbed intact, and IV dosing poses the most conservative test of potential toxicity. However, dosing by this route may not simulate what occurs locally in the GI tract, and thus its relevance to dietary exposure could be questioned. For example, the potential toxicity of antinutrient proteins that interfere with protein digestion and uptake (protease inhibitors, lectins) may not be manifest in the same way if they were administered intravenously instead of by the oral route. For IV dosing, proteins produced in bacteria would need to be highly purified to remove bacterial/fermentation contaminants (e.g., lipopolysaccharides) that are themselves toxic when administered parenterally.84 If there was evidence of toxicity following IV dosing of the protein, acute oral toxicology studies would still need to be conducted to resolve whether these effects were relevant to dietary exposure. Repeat IV dosing is also not recommended as plant-derived proteins would be recognized as foreign to rodents, leading to the development of neutralizing antibodies in the blood that would confound interpretation of study findings. This phenomenon is well documented for the repeated administration of protein-based pharmaceuticals that are not native to the test species (Chapter 6).
EFSA guidelines for testing the safety of biotechnology-derived crops do not recommend acute high-dose testing for insecticidal proteins or for other nonpesti-cidal proteins.81 Rather, EFSA proposes a case-by-case assessment of the safety of introduced proteins, and if the biological profile/activity of the protein raises questions about safety or the protein is considered to be "novel," then a 28-day feeding study with the protein is recommended. This recommendation is appropriate for certain classes of potentially toxic proteins such as lectins or protease inhibitors whose toxicity is manifest after a short-term feeding study.85-86 The characteristics that define an introduced protein as novel have not been elaborated and are best determined on a case-by-case assessment.
It may not be possible to carry out repeat-dosing studies for certain membrane-bound enzymes if they are considered to be novel. Purification and isolation of certain membrane-bound enzymes can lead to their immediate inactivation as membrane lipids and the cofactors needed for catalytic function of the enzyme are removed during purification.87 As a practical matter, there could be negligible dietary exposure to functionally active membrane-bound enzymes in foods if solvent extraction and heat processing (e.g., foods derived from soybeans) results in their inactivation. This may obviate the need for confirmatory safety testing of proteins in animals, given the negligible potential for human and animal dietary exposure.
When an introduced protein is functionally or structurally related to proteins that are toxic to mammals (AFP example, Chapter 10), then an acute high-dose toxicity study may not be sufficient to confirm safety. Other hypothesis-driven studies (based on knowledge of the protein's mode of action) may be necessary, as outlined for the
AFP example. These studies could include a 28-day dietary study with the purified protein in rodents, assuming it could be prepared in sufficient quantities to test.
Not all introduced proteins have pesticidal properties, as some impart other desired traits into crops such as herbicide tolerance, virus resistance, improvements in nutrient content, etc. Often these proteins are enzymes that catalyze specific biochemical reactions. Based on their known mode of action, specificity, lack of functional or structural similarity to protein toxins, digestibility, history of safe use, etc., the weight of evidence would suggest these proteins would not raise food safety concerns. However, in certain countries outside the United States or Europe, regulators have requested high-dose acute studies to provide further confirmation of safety, and proteins that have been so tested are also listed in Table 11.1 (see also Chapter 10). As with the case of PIPs, there has been no evidence to date of adverse effects in mice dosed with high levels of nonpesticidal proteins.
Proteins introduced into biotechnology-derived crops are also components of grain or seed that are formulated into diets and fed to rats for approximately 90 days to confirm the lack of any unintended effects in the biotech crop. Thus, their safety is tested as a component of the grain/seed fed to rats. Other studies, such as molecular characterization of the gene insert, the nutrient/antinutrient composition of food/ feed, the phenotypic and agronomic characteristics of the plant grown in different environmental conditions, and animal performance studies with feed will also have been carried out to assess the potential for unintended effects.
The study design for a 90-day rat feeding study is adapted from OECD 408 guidelines for subchronic studies that include measurement a comprehensive battery of toxicology parameters. Commercial rodent diets used by toxicology testing facilities often include processed soybean meal and corn meal in diet formulations as a source of dietary protein. When new biotechnology-derived corn or soybean crops are developed, they can be incorporated into commercial rodent diets to substitute for conventional corn grain or processed soy meal, and their safety can be assessed. Since the rats are fed levels of corn grain approximately 100 times higher than humans would consume in Europe (assumes conservatively that 100% of the corn grain is derived from the biotechnology-derived crop), these studies can provide confirmation of an acceptable safety margin for the biotechnology-derived crops including the introduced protein(s). If triggered, for example, by results from compositional analysis or differences in phenotypic or agronomic performance, subchronic feeding studies may be conducted to determine whether the biotechnology-derived food is "as safe as" conventional, nonbiotech comparators in accordance with the general principles of substantial equivalence.88-90
Subchronic feeding studies are often required to obtain registration of the biotechnology-derived crop in the EU even though the aforementioned triggers did not occur. It was recently acknowledged in a draft EFSA guideline91 that "In the situation where molecular, compositional, phenotypic and agronomic analysis have demonstrated equivalence between the GM plant derived foods/feed and their near isogenic counterpart, except for the inserted trait(s), and do not indicate the occurrence of unintended effects, the performance of 90-day feeding trials with rodents or with target animal species would be considered to add little if anything to the overall safety assessment. ... These studies did not show any indication for the occurrence of unintended effects." This has been demonstrated in 90-day rat studies conducted to date, some of which have been published in peer-reviewed journals.92-96
Was this article helpful?