Although the vast majority of proteins that are present in our diet do not pose any hazard to human health, a small number of proteins are toxins. Proteins that are toxic usually act via acute mechanisms almost immediately upon consumption.42,43 Hence, evaluation of protein toxicity through acute administration of a single high dose of the protein is considered to be an appropriate test. An additional advantage of the oral gavage (in comparison, for example, with intravenous administration) is that during gavage the protein is subjected to digestion in the gastrointestinal tract as it would when it is present in the food source.
Insect-protected plants expressing insecticidal Cry proteins from B. thuringiensis were among the first commercialized biotechnology-derived crops. Because the Cry proteins are present in B. thuringiensis-based microbial pesticides, which were tested for their toxicity in high-dose, acute gavage studies, the U.S. Environmental Protection Agency (EPA) requested a similar evaluation for the Cry proteins expressed in genetically modified crops.4445 Subsequently, the industry has undertaken the acute mouse gavage with nonpesticidal proteins to support regulatory approval of biotech crops outside the United States, although U.S. and European regulatory agencies do not require this study.
The EPA requires the high-dose, acute oral gavage study to assess the potential hazards of pesticidal proteins to nontarget organisms such as mammals and to establish the no-observed-adverse-effect-level (NOAEL). The NOAEL is the dose that causes no adverse effects in test animals and is used to estimate a safe level of exposure for humans to the food containing the introduced protein, or margin of exposure. The margin of exposure is defined as a ratio of the NOAEL to daily dietary exposure to the transgenic protein, which takes into consideration the quantity of food crop consumed on a daily basis by humans and livestock, and the level of protein expressed in edible parts of the crop. The higher the calculated margins of exposure, the less risk to human and animal health would be associated with dietary exposure to food and feed products containing the transgenic protein. Therefore, a single high dose [g/kg body weight (BW)] has been typically used for pesticidal proteins, the actual dose delivered being influenced by the solubility of the protein in the dosing solution.
Pesticidal proteins such as Cry proteins are 8-endotoxins that bind to specific receptors in the insect's midgut apical microvillar membranes, forming lytic pores and, thus, lyse epithelial cells leading to the death of the target insect.46 Although receptors for these proteins are not present in mammals, the toxic mechanism of action triggers testing of these proteins at very high g/kg BW dosages, providing margin of exposures at orders of magnitude (103 to 106) times higher than human or farm animal dietary exposures. In the case of nonpesticidal proteins (e.g., CP4 EPSPS), which have a well-understood and -described mode of action, a long history of safe consumption, and have demonstrated a rapid digestion with pepsin in SGF, the hazard to human health is extremely low and, therefore, acute toxicity testing is not normally needed. Nonetheless, toxicity evaluation is routinely performed for such proteins as well.
The acute oral toxicity test in mice is a short-term study (~14 days). On the first study day, mice are weighed, fasted for two to three hours, and reweighed prior to dosing. Mice used for the study weigh, on average, approximately 30 g (0.030 kg). Protein dosing solutions are administered at volumes up to 33.3 ml/kg BW or approximately 1 ml/mouse. Typically, protein dosing solutions are administered to groups of 5 to 10 mice/sex at a single-dose level. A negative control group is included where mice are gavaged with an equivalent concentration and dose of a nontoxic protein such as bovine serum albumin (BSA). A vehicle control dose [i.e., the buffer used to formulate the test and control (BSA) protein doses] is also included in the study to make sure that no toxicity is associated with the buffer used for formulation. Animals are than returned to ad libitum feeding after dosing. Body weights are also recorded on Days 7 and 14 and food consumption is measured accordingly. Detailed clinical observations are taken a minimum of two times on Day 0 (post-dose) and daily thereafter (Days 1-14). Clinical observations typically include changes in skin and fur, eyes and mucous membranes, respiratory system, circulatory system, autonomic and central systems (including tremors and convulsions), changes in level of activity, gait and posture, reactivity to handling or sensory stimuli, altered strength, and stereotypes or bizarre behavior. A general health/mortality check is performed twice daily. After two weeks, animals are sacrificed and a gross necropsy conducted. For the gross necropsy, body cavities (cranial, thoracic, abdominal, and pelvic) are opened and examined. Tissues harvested at necropsy are stored for post-study evaluation if needed.
Ideally, the formulated and administered protein doses should undergo minimal loss of purity and functional activity during the time course of the experiment. Samples of the dosing solutions are taken prior to dosing ("pre-dose") and following dosing ("post-dose") and analyzed for total protein concentration and functional activity. Additionally, the doses should be homogenous suspensions, if not demonstrated to be true solutions. Samples of the test and control protein doses are taken from the top, middle, and bottom of the reservoir containing the dosing solutions while stirring so that homogeneity of the doses can be subsequently confirmed by demonstrating equal total protein concentration in these samples. The final dose level (mg of protein/kg BW) is calculated based on total protein concentration (mg/ml), corrected for purity, and multiplied by the dosing rate, which may be up to 33.3 ml/kg BW.
Some of the problems unique to dosing proteins by gavage are due to limitations in protein solubility, lack of protein stability, lack of available toxicity data for buffer components, and lack of available assays demonstrating functional activity. If the target dose level for a pesticidal protein were 5000 mg/kg BW, it would translate to a total protein concentration of 150 mg/ml of dosing solution (assuming 100% purity). Very few proteins are soluble at this concentration. Therefore, proteins are often dosed as suspensions. Even as suspensions, this level of protein concentration may be unattainable. A split dose approach has been employed to circumvent this issue, where two doses are administered on a single day, spaced four hours apart, to a single mouse.
A second issue is lack of toxicity data for many biological buffers and additives that may be important for protein activity or stability. Toxicology data are unavailable for reducing agents [e.g., dithiothreitol (DTT)] and protease inhibitors. Therefore, protease inhibitors are avoided even though these components might be crucial to protein stability, and cysteine or reduced glutathione are substituted as reducing agents for DTT and 2-mercaptoethanol. A large database of acute mouse toxicity data has now been generated for both pesticidal and nonpesticidal proteins (Table 10.1). No evidence of toxicity for either type of proteins has been observed when tested at hundreds- and thousands-fold safety margins.
It continues to make sense to test the acute oral toxicity of pesticidal proteins or proteins with an unknown mode of action and with no history of documented human consumption. However, the value of toxicity testing should be reconsidered when proteins have a long history of safe use, a well-understood mode of action, are not structurally or functionally related to known protein toxins or pharmacologically active proteins, demonstrate rapid digestion in in vitro assays, and are expressed at
Was this article helpful?