Jump to main content.

Data Evaluation Record - Safety Assessment MRID NO: 44714001

Date:  7/15/99

Reviewed by: Michael T. Watson, Ph.D., Plant Pathologist

Secondary Reviewer: John L. Kough, Ph.D., Biologist

STUDY TYPE: Safety Assessment

MRID NO: 44714001

TEST MATERIAL: StarLinkJ Genetically Modified Corn

PROJECT NO:

SPONSOR:  AgrEvo USA Company, Wilmington, DE

TESTING FACILITY: AgrEvo USA Company, Wilmington, DE

TITLE OF REPORT: Safety Assessment of StarlinkJ Genetically Modified Corn Containing the Truncated Bt Insecticidal Protein Cry9C for Human Food Use

AUTHOR: Sally Van Wert, Ph.D.

STUDY COMPLETED: November 20, 1998

CONCLUSION: This data package provides a summary of data and information obtained by AgrEvo regarding the characterization and safety of the Cry9C protein, and the Cry9C-expressing StarLinkJ corn. The data and other information provided does not provide a conclusive argument about the allergenicity potential of the Cry9C protein. The protein does share some characteristics with known allergens, however, it lacks others. Further data/justification should be provided as indicated in the review of each data package contained in this submission, and referenced in this review. Based upon the information provided in this summary, no definitive decision can be reached regarding the allergenicity potential of the Cry9C protein.

CLASSIFICATION: SUPPLEMENTAL. This report is a summary of data and other information submitted by AgrEvo and does not provide additional information not found in the data packages submitted in support of this application.

GOOD LABORATORY PRACTICE: This study is not subject to Good Laboratory Practice standards.

  1. Summary

    2. Background

    StarLinkJ Corn has been genetically modified to express both the Cry9C and phosphinothricin acetyltransferase (PAT) proteins. The Cry9C protein is expressed via a modified cry9C gene isolated from Bacillus thuringiensis subsp. Tolworthi. The PAT protein (confers resistance to glufosinate ammonium herbicides) is encoded for by the bar gene from Streptomyces hygroscopicus. The PAT protein has been previously tested and is not the primary concern of this report. The focus of this report is the safety of the Cry9C protein, as it is expressed in the StarLinkJ corn plants.

    The Cry9C protein has 1157 amino acids with the insecticidal fragment contained between amino acids 44 and 658. The truncated endotoxin consists of approximately 626 amino acids encoding for a crystal protein of approximately 70 kDa. In Cry9C expressing plants, the protein is not found in its normal crystalline form, but as a free in plant cells. The mode of action of this protein, as with other Bt proteins, is mediated by receptor binding in the insect gut.

    However, unlike other Bt proteins used in transgenic plants, Cry9C is significantly more stable. The protein was stable at temperatures up to 90E C and remained intact following 4 hours in simulated mammalian gastric juice (in vitro). It is these two characteristics which have raised concerns regarding the safety, specifically the allergenicity of this protein if it was part of the human diet.

    This submission consists of reviews of tests performed for or by AgrEvo on the Cry9C protein and/or the StarLinkJ corn. In addition, there are literature citations of studies relating to these topics and comments/assessments made by scientists with knowledge in these subject areas. This safety assessment

    1. Characterization of the parent corn


    2. The transformation process


    3. Characterization of the gene product (i.e. Cry9C protein)


    4. Characterization of the genetically modified (GM) corn

    Through these four subject areas, AgrEvo attempts to show that their StarLink J Corn line is "substantially equivalent in all essential aspects" to its unmodified parent. The information and data provided includes a description of the protein and summaries of: human health effects testing; insecticidal activity; in vitro binding of the Cry9C protein; potential allergenicity; protein stability; immune response to the Cry9C protein; and levels of expression of the protein in the CBH-351 line of corn.

    Summary of Data

    1. Characterization of the Transformation Process for StarLinkJ Corn (CBH-351)
      1. Source of cry9C and bar Genes - The CBH-351 transformation event contains three foreign genes: the cry9C gene modified from the insecticidal crystal gene of B. thuringiensis subsp. tolworthi; the bar gene form Streptomyces hygroscopicus; and the bla marker gene (confers resistance to ampicillin).
        2. >

      2. DNA Insert, Sequence, and Purity - [CBI Deleted] A series of Southern analysis have shown that one copy of the cry9C gene and four copies of the bar gene have been inserted at a single site in the corn genome.


      3. Pleiotrophic Effects - Expression of the cry9C gene occurs in all major tissues of the corn plant susceptible to European corn borer. This was accomplished by linking the gene to a constitutive and high expression promoter.


    2. Production and Safety Characterization of Cry9C Protein:
      1. Protein Structure - The bacterial Cry9C protein has 1157 amino acids, with the insecticidal fragment contained between amino acids 44 and 658, encoding for a crystal protein of 70 kDa. The Cry9C protein in planta is free in the plant cells, not in crystalline form.


      2. Insecticidal Activity of Cry9C Protein in Mammals - Mode of Action - The mode of action of Cry9C, as with other Bt proteins, is mediated via receptor binding in the gut in insects. Further testing was performed to determine how this correlates with any Cry9C protein activity in mammals. This is discussed further in Section C below.


      3. In vivo Binding of Cry9C Protein to Insect GI-Tract Tissues - Biological Target Specificity - Previous studies with the European corn borer have shown that the crystal Cry9C protein recognizes a receptor, different from that recognized by the Cry1Ab5 protein. It has also been shown that elimination of the trypsin cleavage site, which is responsible for the degradation from a toxic 70 kDa protein to the non-toxic 55 kDa protein (resistant to protease digestion), did not increase the toxicity in the target insect.


    3. In vitro Binding Potential of the Cry9C Protein to the GI-Tract Tissues of Mammals:
      1. In vitro - A study included in this same data package (MRID# 447343-01) investigated the binding of the Cry9C protein to the brush border membrane vesicles (BBMVs) from the midgut of European Corn borer and the BBMVs from mouse intestinal preparations. In this submission, AgrEvo indicates that they believe that addition of labeled Cry9C resulted in displacement of the labeled Cry9C in the insect BBMVs, but did not result in displacement of the labeled Cry9C in the mouse BBMVs. Therefore, according to AgrEvo, this would indicated that Cry9C binds specifically and saturably to the European Corn Borer BBMVs and not to the mouse intestinal BBMVs. However, the figures included in this submission are of very poor quality and it is not possible for EPA to reach the same conclusion. If adequate pictures are available, further justification for the belief by AgrEvo that displacement is occurring in this case should also be provided. Supplementary figures have been requested (as described in the review of MRID# 44734301) for further analysis of these results.


      2. In vivo - A 30-day mouse toxicity study was included in this submission under MRID# 44734303


    4. Evaluation of the Potential for Mammalian Toxicity with the Cry9C Protein in Mice:
      1. 1. Acute Oral Toxicity - The acute oral toxicity of the Cry9C protein has been previously assessed under MRID# 44258107 (Kough to Mendelsohn, 1/6/98). The data from this study indicated that there were some clinical signs due to the dosing, but no deaths at a dose above the limit dose of 5000 mg/kg bodyweight.


      2. 2. Acute Intravenous Toxicity - An acute intravenous toxicity study was also included in this submission under MRID# 44734302. The data in this study indicated that mice dosed at 0.3 mg/kg body weight did not produce any apparent adverse effects in the mice tested.


    5. Potential for Allergenicity:
      1. Molecular Weight - The Cry9C protein has a molecular weight of 68.7 kDa and is potentially glycosylated post-transcriptionally (no glycosylation was measured in either the plant or bacterial produced Cry9C). According to AgrEvo, the majority of food allergens have a molecular weight between 10 and 40 kDa and are often glycosylated.


      2. Sequence Homology - The optimal number of amino acids needed to elicit an immunological response appears to be between 8 and 12. No match between any 8 amino acid sequences in the Cry9C protein and any known allergenic proteins in the SWISS protein database, (previously reviewed: MRID# 44384404 - Kough to Mendelsohn, 4/8/98) or the other available databases was found.


      3. Stability to Digestion - Stability to digestion is another characterization common to known allergens. Purified Cry9C has been tested in vitro in simulated mammalian gastric fluid (previously reviewed: MRID# 44258108 - Kough to Mendelsohn, 1/6/98). The samples of lyophilized Cry9C protein expressed in corn showed no signs of protein disintegration when subjected to simulated human gastric conditions. These digestions were done either with or without pepsin in low pH buffer and assayed by Western blot on samples taken up to 4 hours after exposure.


      4. Stability to Heat and Processing - The Cry9C protein was shown to be stable at a temperature of 90E C for 10 minutes, without altering the toxicity to the target insect (previously reviewed: MRID# 44258108 - Kough to Mendelsohn, 1/6/98).


    6. Immunological Response
      1. 1. Animal Model - The Brown Norway Rat has been chosen by AgrEvo as a surrogate for testing the potential immune response in humans to Cry9C. This animal has been identified as a high IgE responder. The production of an IgE antibody response to a food antigen is part of an atopic syndrome. This rat has been used in a study included in this submission (MRID# 44714002).


      2. 2. In Humans - AgrEvo indicates that many workers (1990) have been involved in seed production, planting, harvesting, and processing the corn expressing Cry9C over a period of time (since 1996). No cases of sensitization or allergenic response have been reported for these workers. In addition, as previously reviewed (MRID# 44384404 - Kough to Mendelsohn, 4/8/98), 21 sera samples from suspected corn-sensitive individuals were screened with corn seed extracts from Bt Cry9C corn (CBH-351). The sera were assayed for specific IgE to aqueous wild-type or transgenic corn allergens by Radio Allergo Sorbent Test (RAST). All of the sera tested positive in the RAST assay by having a $3% reactivity. The transgenic and wild-type aqueous extracts were not obviously different in responsiveness for individuals and a t-test of the RAST % reactivity did not reveal any significant differences.


    7. Levels of Expression

      8. Included in this submission is a study which investigates the levels of Cry9C expression in corn, and various processed commodities (MRID# 44734303). Unfortunately, the control corn line grown in the same location as the test line showed positive signals for both the Cry9C and PAT proteins. Because of this contamination, the results of this study are questionable. However, the study indicates here that the typical values of Cry9C in corn ears are about 0.5% of total protein.

    8. Bioavailability

      9. The bioavailability of the Cry9C protein was examined (H.P.M. Noteborn, 1998) in a single dose gavage study in the rat, where blood samples were removed over an 8 hour period. Both ELISA and Western Blot analysis were used for detection of the Cry9C protein. The author of the study concluded that very small traces of Cry9C-like material was detected in the blood at the top dose. The report indicated that of the 298 mg/kg body weight dose, between 0.0002 and 0.0006% was absorbed (these values were either on or below the reported limit of detection of the assay). The identity of the Cry9C-like material could not be confirmed by Western Blot analysis.

      Also included in this study were tests designed to confirm that the Cry9C protein is degraded in the intestinal system of the rat to a protein of 55 kDa and that gut bacteria may play a role in part of this degradation. AgrEvo indicates in this report that this result casts doubt over the importance of the in vitro (i.e., Cry9C stability in vitro) results and provides strong evidence for in vivo degradation of the Cry9C protein. However, BPPD does not necessarily agree with this interpretation. It may be true that the protein is broken down from a 68 kDa protein to a 55 kDa protein, but there is no additional evidence that the protein is further degraded. The apparent stability of the 55 kDa protein is not addressed, and therefore continues to pose an allergenicity concern.

    9. Wholesomeness

      10. Poultry Feeding Study/Wholesomeness - An assessment of the wholesomeness of the GM corn versus control corn was conducted in chickens (S. Leeson, 1998). Chickens were separated into groups and received a diet of either the control corn or corn containing Cry9C at 65% (w/w). The chickens were weighed weekly and observed for clinical signs for up to 42 days. Results showed no adverse effects on feeding, growth, or clinical condition of the animals.

      Irritancy - AgrEvo cites two studies (D.A. Douds, 1997; H.P.M. Noteborn, 1998) which indicated that mice in an acute oral study and a 30-day mouse repeat dose study did not show any evidence of irritation in their gastrointestinal tracts.

    10. Consumer Safety:

      11. AgrEvo indicates that there are no known reports of adverse incidents to man exposed to the Cry9C protein in seed, plant or bacterial form from occupational exposure. The company further attempts to calculate a Theoretical Maximum Daily Intake (TMDI) for the Cry9C protein. The calculation as based on the possible content of the Cry9C protein in various commodities and on the maximum measured levels of Cry9C in processed grain. The global TMDI was 0.00923 and the European TMDI was 0.00193 mg/person/day. The acceptable daily intake or reference dose figure of 0.3 mg/kg/day was derived from the NOEL or margin of exposure (i.e., 20 mg/kg/day) in the 30-day mouse study. The figure obtained is subject to the addition of safety factor of 100.

  2. DISCUSSION

    3. The information provided in this submission is intended to support the belief by AgrEvo that the Cry9C protein, and the StarLinkJ corn plants expressing this protein, do not pose a significant risk to human health. Some of the data and information provided by AgrEvo is compelling, and supportive of the belief of not significant risk. However, there is at least an equal amount of data and information that is either inconclusive, or indicates that Cry9C exhibits some characteristics of known allergens.

    Data which is supportive of the non-allergenic nature of the protein include its size, its lack of amino acid sequence relatedness to other known allergenic proteins and its level of expression/titer in transgenic plants. Although there are no absolute characteristics which are definitive for allergenicity potential, taken by themselves, these characteristics do not necessarily raise concern about the protein. However, the stability of the protein at high heat (90E C) and in simulated gastric juice, do raise concerns about the protein.

    Much of the remaining information provided in this submission is inconclusive. The data from the field trials (MRID# 44734304), where the control samples showed the presence of the Cry9C and PAT proteins, makes this study supplemental at best. Because no explanation was provided for the problems with the control, it casts some doubt on the results of the test samples and the subsequent control samples used for the validation study. The protein binding study (MRID# 44734301) is also supplemental because the figures included in the submission are not adequate to make a determination about the results of the study. The acute intravenous toxicity study (MRID# 44734302) is an acceptable study, and indicates that there is no apparent toxicity from the protein dosed at 0.3 mg/kg body weight. However, this only provides further support to the conclusion reached in review of previous toxicity studies, that the protein is apparently not toxic.

    Given the acceptable status of the intravenous study, and assuming the other two studies prove to be acceptable, they do not support a conclusive case that the Cry9C protein is not a potential allergen. There is not overwhelming evidence to indicate that Cry9C is a potential allergen, but there is not enough evidence to indicate that it is not. Included in this submission is: "An Expert Assessment of the Allergenic Potential of StarLinkJ Corn", written by Dr. Andrew Cockburn. Dr. Cockburn concludes that, "StarLinkJ corn is safe in all aspect as the unmodified corn used today". Further, a study by BIBRA International entitled: Development of New Methods of Safety Evaluation of Transgenic Food Crops, is also enclosed. The objective of this study is to investigate the potential immunogenicity and allergenicity of recombinant novel food proteins in transgenic tomatoes. The study indicates that the brown Norway rat, as used by AgrEvo for the Cry9C protein, is a suitable model for investigation of the allergenic components of foods. However, a letter also included in this submission, written by Dr. Samuel B. Lehrer indicates that he is skeptical about the applicability of this rat model in determining the allergenic potential of the Cry9C protein.

    As with much of the data and information submitted in this package, it is not possible to reach a definite conclusion regarding the allergenic potential of the Cry9C protein. Some of the data do not support the lack of allergenic potential, while other data indicates the possibility of properties similar to other allergens. Other data is basically supplementary, and does not directly address the protein, or its allergenic potential directly. As can be seen from the enclosed letter and reviews, even expert analysis of this protein differs. Based upon these factors, it is not possible for BPPD to determine that there is a lack of allergenic potential from Cry9C based upon the available information. Further data and clarification must be provided to aide in such a determination.

Publications | Glossary | A-Z Index | Jobs

Local Navigation

Jump to main content.