Grantee Research Project Results
Final Report: Comparison of Gnotobiotic and Conventional Mice for Predicting the Allergenic Potential of Proteins Introduced into Genetically Engineered Plants
EPA Grant Number: R834824Title: Comparison of Gnotobiotic and Conventional Mice for Predicting the Allergenic Potential of Proteins Introduced into Genetically Engineered Plants
Investigators: Baumert, Joseph L , Goodman, Richard E. , Peterson, Daniel H
Institution: University of Nebraska at Lincoln
EPA Project Officer: Aja, Hayley
Project Period: September 15, 2010 through September 14, 2013 (Extended to September 14, 2014)
Project Amount: $423,546
RFA: Approaches to Assessing Potential Food Allergy from Genetically Engineered Plants (2009) RFA Text | Recipients Lists
Research Category: Human Health
Objective:
The research focused on the development of a more reliable, practical, and predictive animal model that can be used to evaluate the allergenic potential of proteins introduced into genetically engineered plants. The primary objective of this research project was to evaluate sensitization responses in germ-free (GF) mice and mice having a conventional (CONV) intestinal microbiota to orally presented purified proteins (potent allergen, peanut Ara h 2; moderate allergen, bovine beta-lactoglobulin [BLG] and nonallergenic soybean lipoxygenase [LOX]).
Summary/Accomplishments (Outputs/Outcomes):
Sensitization by the intraperitoneal (IP) route and challenge by the IP route was deemed superior to intragastric (IG) sensitization and challenge, based on the reduced variance in responses. Additionally, from a practical standpoint the IP route is time and resource efficient. Ten-fold less protein is required for a complete mouse experiment using IP sensitization and challenge when compared to IG sensitization and challenge. Purifying a protein is labor and cost intensive, and obtaining several grams of high-purity protein is not always possible; the protein-efficient IP sensitization and challenge protocol could be utilized for risk assessment in situations where enough protein for IG route is not available. Although the differences between GF and CONV mice sensitized with BLG were not as significant as we hypothesized, the variance was reduced in GF mice compared to CONV, which is why we utilized GF mice when comparing different proteins of varying allergenicity in the model system.
Characterizing both clinical and serological responses is important for evaluating the allergic response in an animal model. When looking at the various markers it seems clear that this model is able to differentiate between known allergens based on potency. As pure protein challenges are not performed in humans, some informed assumptions are necessary. We assume Ara h 2 is more potent than BLG because more severe reactions are observed to peanut than milk. These specific proteins are indicated as allergens in humans based on sera testing for IgE binding, protein characteristics, and skin prick tests. We assume LOX is a nonallergen based on a lack of reported IgE binding to LOX from soybean-allergic patients. That being said, the observed responses correlate with the overall allergenicity response to these proteins in humans, which is essential if this model will be useful in assessing the allergenic risk of novel proteins in the future.
This animal model is able to differentiate between purified potent and nonallergens. As far as ranking the two potent allergens in this study, hypothermia was the only indicator that Ara h 2 was more potent than BLG. The other indicators—mMCP-1, sIgE positivity, and clinical scores—were similar between the two allergens. The responses to both known allergens, Ara h 2 and BLG, were significantly greater than responses observed in mice sensitized and challenged with the nonallergen, LOX.
The results indicate this animal model was able to reliably differentiate proteins based on inherent allergenicity as strong responses to potent allergens and weak responses to a non-allergen were observed.
Conclusions:
A robust, predictive animal model of food allergy requires a strong response to potent allergens and little or no response to nonallergens, as seen in this study. Additionally, a robust and predictive risk assessment tool must be reproducible from experiment to experiment or laboratory to laboratory to have value for the industry or regulators. Our mouse model has proven to be reproducible in our facility, where we utilized large sample sizes to carefully evaluate the statistical variance of the risk assessment model. Each treatment group in our study utilized eight to nine treatment mice and five control mice per protein per treatment. In addition, we included independent replication of each treatment group in order to carefully evaluate the statistical variance of our predictive model. If this model were to be widely adopted, the absence of the gut microbiota in GF mice likely will allow for a level of reproducibility between laboratories not previously seen. By using GF mice, we believe we have taken out an important environmental variable that may be responsible for the high variance we observed when using conventional mice.
Journal Articles:
No journal articles submitted with this report: View all 10 publications for this projectSupplemental Keywords:
anaphylaxis, clinical symptoms, conventional, food allergy, germ-free, gut microbiota, IgE, mast cell protease, murine model, purified dietary allergen challenge, risk assessmentProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.