Interpretation of Safety Assessment Data for GM Foods

General Information
  1. Introduction

    Based on the principles developed by international organizations like the World Health Organization (WHO), the Food and Agriculture Organization (FAO), and the Codex Alimentarius Commission (CAC) jointly established by FAO/WHO, the Organization of Economic and Co-operation and Development (OECD) and the US Food & Drug Administration (FDA), the aim of safety assessment of biotechnologically derived food products is to provide assurance, using the best available scientific knowledge that these foods are as safe to consume as their conventional counterpart and therefore do not cause appreciable harm when prepared, used and/or eaten according to their intended use. In general terms, to determine the safety of genetically modified foods, some of the questions that must be answered in the assessment process include:

    • Do the donor and recipient organisms have a history of safe use?
    • Are the new substances produced e.g. proteins safe to consume?
    • Have potential allergens been introduced into or increased in the food?
    • Are there changes in the content of other important substances e.g. toxicants, antinutrients?
    • Has the composition and nutritional value changed?
    • In what forms will the food or food products derived from it be consumed?
    • Do the newly introduced substances survive processing, shipment, storage, and other preparation?
    • What is the expected human dietary exposure?
    • If an antibiotic resistance or other selectable marker is present, is it safe?

    There are also questions to be answered about characterizing the transformation at the molecular genetic level e.g. where is the gene construct incorporated in the plant genome, has the sequence been characterized, are unexpected open reading frames present, number of copies of the gene, stability of expression etc.

    Based on the Codex Guidelines for the safety assessment of foods derived from rDNA plants that was covered in a different section, a case study example – Roundup Ready soybean (RR soy), will be used to explain key data that needs to be availed as part of the food safety assessment application dossier and how to interpret the data. The case study information here includes excerpts from an application for food safety assessment dossier submitted by Monsanto company to regulatory authorities in Canada, USA and the UK and made available courtesy of the FAO.

    Case Study – Roundup Ready Soybean (RR soy)

    Monsanto Company has developed RR soy (event 40-3-2) varieties that confer tolerance to glyphosate, the active ingredient in Roundup herbicide, a herbicide used by farmers to kill weeds. The CP4 enolpyruvylshikimate-3-phosphate synthase (EPSPS) protein is an important enzyme involved in the biosynthesis of aromatic amino acids in plants and microorganisms. (EPSP synthase is not present in animals and therefore glyphosate has a high level of safety for consumers) Inhibition of this enzyme by glyphosate leads to a deficiency in the production of aromatic amino acids and lack of growth in plants. RR soy (event 40-3-2) was produced by introduction of a cp4 epsps gene from a naturally occurring soil bacterium (Agrobacterium) using particle-acceleration transformation into the soybean genome. The cp4 epsps gene encodes for a version of the ESPS enzyme that is tolerant to glyphosate, meaning that Agrobacterium is not killed by the herbicide.

    The RR soybean plant therefore produces two EPSPS enzymes: the soybean version already present in the plant, and the bacterial version added in the genetic modification. When the herbicide is applied, the soybean EPSPS enzyme is blocked and cannot function. However, the plant survives because the bacterial EPSPS enzyme that has been transferred into the plant remains active, allowing it to continue to make the necessary amino acids. Thus, the bacterial protein is able to function in the soybean in the same way that it does in the soil bacterium.
  2. Description of the Recombinant-DNA Plant

    A description of the rDNA plant being presented for safety assessment should be provided to identify the crop, the transformation event(s) to be reviewed and the type and purpose of the modification. The description should be sufficient to aid in the understanding of the nature of the food being submitted for safety assessment.

    Case Study

    The RR soy (event 40-3-2) was developed to allow for the use of glyphosate, the active ingredient in the herbicide Roundup, as a weed control option. This genetically engineered soybean line contains a form of the enzyme 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) that allows event 40-3-2 to survive the otherwise lethal application of glyphosate. The EPSPS gene put into event 40-3-2 was isolated from a strain of the common soil bacterium Agrobacterium tumefaciens called CP4. The form of EPSPS enzyme produced by this gene is tolerant to glyphosate.
  3. Description of the Host Plant and Donor Organism and Use as Food

    This section of the data looks at the phenotypic and agronomic description of the plant, history of use of the conventional food including identifying the edible components of the food, food products commonly containing these edible components and processing requirements. May also include information of on how the plant is typically cultivated, transported and stored and whether special processing is required to make the plant safe to eat, and the plants normal role in the diet. As far as the donor organism is concerned, there should be information for its identification and it is particularly important to determine if the donor organism (s) or other closely related members of the family naturally exhibit characteristics of pathogenicity or toxin production, or have other traits that affect human health.

    Case Study

    RR soy has been tested in field trials in the United States, Central and South America, Europe, Central Europe and Canada since 1991. Data collected from trials conducted in over a 3-year period prior to commercialization in the United States demonstrate that RR soy does not differ significantly from conventional soybeans in morphology, seed production (yield), agronomic characteristics such as time to flowering and pod set, or vigor (germination or persistence). In addition, RR soy was monitored for its susceptibility to diseases and insects, and there were no differences observed in disease severity or insect infestations between RR soy and the control plants.
    For RR soy, the host (recipient) organism is the soybean plant and the part of the soy plant that is consumed by humans is the seed — the soybean. Soybeans (soy) can be cooked and consumed whole or processed into many types of foods. Foods made from soy include soy beverages, tofu, soy oil, soy flour and lecithin. Some of the processed products containing soy may include breads, pastries, snack foods, baked products, fried products, edible oil products and special purpose foods such as infant formula.
    The donor organism for RR soy is the common soil bacterium Agrobacterium (species strain CP4), from which the EPSPS gene was derived. Agrobacterium is found everywhere in the environment, so it could be expected to be commonly encountered in foods harvested directly from soil, such as cassava, sweet potatoes etc. As Agrobacterium has been ingested by humans over a long period of time, and there is no evidence that it causes toxicity or allergenicity in humans, it is reasonable to assume that the donor organism is safe and acceptable.
  4. Description and Characterization of the Genetic Modification

    All characteristics of the genetic insert must be known including the identity of the source of genetic material, the nucleotide sequence of the DNA construct being inserted, the number of insertion sites, and the stability of the insertion in the plant genome. The data requirement related to the genetic modifications serve to allow a detailed a detailed understanding of the resulting genetic insertions and their locations in the host plant. A detailed description of the molecular characteristics of the rDNA plant is required in order to demonstrate that the developer has critically analyzed the plant and its products, including all introduced genes and expressed proteins.

    1. Method used for modification

      A detailed description of the method used to insert the new genetic material should be provided.

      Case Study

      RR soy was produced using a particle-acceleration transformation method. In this process, the plant cells are bombarded with microscopic particles of gold or tungsten coated with DNA that contains the new EPSPS gene from the donor organism (Agrobacterium). The aim is to get the new gene across the plant cell wall, so that it can be incorporated into the plant cell’s genetic material.
    2. Function

      To understand how the new gene(s) function in the plants one would need to know the new gene(s) and their products, in this case the enzyme EPSPS, the genetic material that controls how, where and when the new genes are switched on and the genetic material that targets any new proteins to specific parts of the cell.

      Case Study

      RR soy contains a single new (cp4 epsps) gene, which codes for the EPSPS enzyme. In front of the bacterial EPSPS gene in RR soy is a regulatory sequence that tells the plant to turn the gene on, known as E35S promoter. At the other end of the EPSPS gene, another regulatory sequence tells the plant where the EPSPS gene ends, known as NOS 3′. A further regulatory sequence is the chloroplast transit peptide sequence whose function is to tell the plant cell to transport the bacterial EPSPS enzyme into the chloroplast of the cell.
    3. Characterization of the New Gene

      This is detailed information on the arrangement of the new genetic material in the genome of the host organism including how many complete or incomplete copies of the new genetic material are present. It is also useful to compare the DNA sequence of the new genetic material in the modified plant’s genome with that of the original DNA to determine if there are any unexpected changes in the DNA sequence in the plant.

    4. Stability of the Genetic Changes

      The genetic changes in a genetically modified plant must be stable. The new genetic material is considered to have become a stable part of the host genome if they remain the same over several generations of plants produced by conventional breeding. The This means that the newly introduced traits should be shown to pass from one generation to the next in a normal predictable way, following the principles of inheritance.

      Case Study

      A single full copy of the bacterial EPSPS gene plus flanking DNA sequences was present in the genome of RR soy. This gene was intact and of the correct size and sequence.
      Case Study

      The inserts are inherited in the expected Mendelian pattern and the stability of the inserts has been demonstrated by molecular analysis of the 3rd through the 6th generations of event 40-3-2. In addition, RR soy event has been in commercial production globally since 1996 with consistent product performance.
    5. Effects of the new gene on human health

      Concerns have been expressed that new genes, particularly antibiotic resistance genes, in GM foods might be able to transfer to disease-causing bacteria in the human digestive tract. Whether new genetic material in GM foods could transfer to gut bacteria and impact on human health has been considered in detail by the World Health Organization (WHO) and several European expert advisory panels, and in numerous scientific papers published in peer reviewed journals. There is general agreement that current genetic materials in GM foods has passed risk assessments and are not likely to present risks for human health. In addition, no effects on human health have been shown as a result of the consumption of such foods by the general population in the countries where they have been approved (WHO, 2009).

      Two issues here (a) gene transfer to gut microorganisms (b) transfer into the human host (the gene may not be expressed but could incorporated fragments cause mutations?). The possibility of this happening is very low due to the highly unlikely series of events that would be required for plant DNA to transfer genes horizontally into mammalian or bacterial genomes (Chassy, 2002). However, this possibility cannot be completely discounted (Jonas et al., 2001). Nevertheless, in a recent report the European Food Safety Authority has reaffirmed that two commonly used antibiotic resistance marker genes, npt II and aadA, pose no threat to humans or the environment (EFSA, 2009).

      Case Study

      As RR soy contains no antibiotic resistance genes, this is not a concern in this case. The only new gene in RR soy that could potentially be transferred to cells in the human digestive tract is the bacterial EPSPS gene. No impact is expected on human health if this gene were to transfer to bacteria in the human gut, because the EPSPS enzyme in RR soy would function in the same way as the EPSPS enzyme naturally present in the gut bacteria. Finally, there is no evidence that DNA sequences from ingested foods have ever been taken up by human cells and incorporated into human DNA (Chassy, 2002).
  5. Characterization of New Proteins
    1. Nature of the new protein

      The nature and function of any new proteins in the GM food is also examined as part of the assessment process. This information also includes verification that the size of the new protein is as expected, and to quantify how much new protein is present in particular tissues. The presence and level of new proteins in particular components of GM varieties used as food, or in food preparation, may present safety issues. This should be assessed in the parts of GM plants that are actually eaten. It is possible that the new protein is only expressed in non-edible parts of the plant, inactivated, denatured or removed by heat or processing (cooking) or only present at very low levels in the edible part of the plant.

      Case Study

      The CP4 EPSPS protein produced in RR soy is functionally similar to a diverse family of EPSPS proteins typically present in food and feed derived from plant and microbial sources. In the safety assessment, the level of the bacterial EPSPS enzyme in fresh or processed edible parts of soy was found to make up less than 0.1% of the total protein. The EPSPS enzyme was shown to have no activity in the edible parts of the soy, because the enzyme is inactivated by heat during processing of the food.
    2. Potential toxicity of new protein

      In this section, the assessment examines the potential toxicity of any new proteins in the GM food. Information should be provided to ensure that genes coding for known toxins or anti-nutrients present in the donor organisms are not transferred to recombinant-DNA plants that do not normally express those toxic or anti-nutritious characteristics.

      In the case of proteins, the assessment should focus on amino acid sequence similarity between the protein and known protein toxins and anti-nutrients as well as stability to heat or processing and to degradation in appropriate representative gastric and intestinal model systems. Appropriate oral toxicity studies may need to be carried out in cases where the protein present in the food is not similar to proteins that have previously been consumed safely in food.

      Non-protein substances that have not been consumed in food should be assessed on a case-by-case basis depending on the identity and biological function in the plant of the substance and dietary exposure. This may require the isolation of the new substances from the recombinant-DNA plant, or the synthesis or production of the substance from an alternative source, in which case the material should be shown to be biochemically, structurally, and functionally equivalent to that produced in the recombinant-DNA plant.

      Case Study

      The EPSPS protein does not show meaningful amino acid sequence similarity when compared to known protein toxins. Rapid degradation of the CP4 EPSPS protein correlates with limited exposure to the gastrointestinal tract and little likelihood that the protein can produce pharmacological, toxic, or allergenic effects. There were no treatment-related adverse effects in mice when CP4 EPSPS protein was administered orally at dosages of up to 572 mg/kg of body weight. This dose represents a significant—greater than 1,000-fold—safety margin relative to the highest potential human consumption of CP4 EPSPS protein. Finally, there is no history of EPSPS proteins being toxic. The bacterial EPSPS enzyme in the soybean has a similar structure and function to the EPSPS enzyme naturally present in the soybean and other plants forming part of the human food supply. On the basis of the evidence described above, the bacterial EPSPS enzyme in RR soy is unlikely to be toxic.
    3. Potential allergenicity of new protein

      This component of the assessment looks at whether any new protein present in the RR soy is likely to cause an allergic reaction in some people. The stepwise approach for evaluating the allergenic potential of the genetically modified food examines the following parameters:

      • Source of the protein – Genes derived from known allergenic sources are assumed to encode an allergen unless scientific evidence demonstrates otherwise. Also attention should be paid to the choice of the expression host, because the post-translational modifications allowed by different hosts may impact on the allergenic potential of the protein.
      • Amino acid sequence homology – The purpose is to assess the extent to which a newly expressed protein is similar in structure to known allergens to assess the allergenic potential of the protein.
      • Pepsin resistance – There exists a correlation between resistance to digestion by pepsin and allergenic potential and therefore the resistance of a protein to degradation in the presence of pepsin under appropriate conditions indicates further analysis should be undertaken to determine whether the expressed protein is potentially allergenic.
      • Specific serum screening – Proteins that originate from a source known to be allergenic, or that have sequence homology with a known allergen, immunological assays should be performed where sera are available. There should be additional testing such as the possible use of skin test and ex vivo protocols even if the in vitro immunoassay comes out negative.
      Case Study

      The amino acid sequence of the CP4 EPSPS protein was compared to protein sequences associated with allergy and coeliac disease and shown to have no meaningful amino acid sequence similarity with any of the known allergens. The protein does not represent a previously described allergen and does not share potentially immunologically relevant amino acid sequence segments or structure with a known allergen. The in vitro assessment of the CP4 EPSPS protein digestibility indicates that the protein, like other food-derived proteins, is very labile to digestion when compared to many clinically important food allergens.

      EPSPS protein is present at low levels, approximately 0.08% of the total protein, in whole RR soy seed. Most food allergens are present as major protein components in the specific food, in amounts ranging from 1% up to 80% of total protein. An assessment of the endogenous allergens in conventional and RR soy was made using sera from patients confirmed to be sensitive to soybean protein and the study demonstrated that the introduction of the EPSPS protein did not cause any discernible changes, either qualitatively or quantitatively, in the composition of the allergenic proteins endogenous to soybean. It has been shown that the processing steps used in the production of soybean oil, one of the main sources of soybean in the human diet, reduce the vast majority of protein such that refined soybean oil did not trigger allergenic reactions in humans who were sensitive to soybean. Based on the evidence described above, the bacterial EPSPS enzyme in RR soy is unlikely to be allergenic.
  6. Compositional Analyses of Key Components

    The GM crop and the conventional variety from which the GM crop was produced (or a variety as close as possible to the GM crop) are grown at different locations that represent areas where the GM crop might be grown in the future. The edible part, in this case the seed, is harvested and tested to determine its composition. The levels of nutrients and other components of the GM crop are compared to the conventional food and also compared to levels reported in the scientific literature for other varieties of the same crop. Any significant differences between the GM and the conventional crop are then assessed for potential adverse health effects.

    1. Nutrient analysis

      Typical nutrient analyses are:

      • Proximate composition – refers to the levels of ash, moisture, protein, fat, and fiber.
      • Amino acid analysis
      • Fatty acid analysis
      • Carbohydrate analysis
      • Vitamin and mineral analysis

      Other compounds present in particular foods may also be measured if they are likely to have a significant impact in the overall diet. For example, the assessment would consider isoflavones in soybeans. Soybeans naturally contain isoflavones and their ingestion has been linked to a number of biochemical effects in mammalian specie including estrogenic and hypocholesterolemic activities (Wang et al., 1990) and have also been reported to contribute deleterious effects on livestock animals fed on soybean meal (Setchell et al., 1987)

      Levels of antinutrients – This test is carried out to check that the genetic modification has not significantly increased levels of any known naturally occurring antinutrients in the food above the natural range found in the conventional food. Processing of the foods must also be taken into account, because this may inactivate any antinutrients in the unprocessed food. Antinutrients found in conventional foods include trypsin inhibitors and phytic acid.

      Rounded Rectangle: Case Study

      The only significant antinutrients known to occur naturally in soybeans are soybean lectin, trypsin inhibitor and phytate. Soybean lectin and trypsin inhibitor are destroyed during the heat treatments or processing applied to all soy products before they are eaten. Nevertheless, no differences were found in the levels of soybean lectin, trypsin inhibitor and phytate between RR soy and conventional soy. Two other components were also measured: raffinose and stachyose – while not strictly antinutrients, increased levels of these carbohydrates would be considered undesirable because they cause flatulence. No differences were found in the levels of raffinose or stachyose between RR soy and conventional soy.
  7. Nutritional Assessment

    The nutritional adequacy and the ability of a GM food to support typical human growth and well-being should be established. This is usually achieved by understanding the genetic modification and its consequences, and analyzing the composition of the food. If the compositional analysis indicates significant differences in a number of important nutrients or other components, or if there is concern that the bioavailability of key nutrients may be compromised by the genetic changes to the food, then feeding studies in animals can determine whether the food is nutritionally adequate.

    Rounded Rectangle: Case Study

    In the case of RR soy the extent of the compositional and other data were considered sufficient to establish the nutritional adequacy of the food. These studies confirmed that there were no unexpected changes in palatability or wholesomeness of RR soy. In these studies, RR soy and conventional soy were fed to some animals that commonly eat soybeans including groups of rats, chickens and dairy cows for 4–6 weeks. These animal studies showed that RR soy was palatable and able to support typical growth and wellbeing in rats, chickens and dairy cows. They also confirmed the results of studies showing no acute toxicity when the bacterial EPSPS enzyme was given to mice at high doses. Feeding studies with catfish and quail also gave results that were consistent with the studies on rats, chickens and dairy cows.
  8. Other safety issues

    Other safety issues that are important on a case-by-case basis should also be considered. For example in the case of RR soy, the possibility of indirect pesticide residue accumulation due to the herbicide tolerance trait would be investigated. In this case traditional methods of chemical safety testing should be applied.

    Rounded Rectangle: Case Study

    In the case of RR soy, the levels of glyphosate used were established to be within allowable safety limits.
  9. Local Concerns

    Where appropriate, safety risk assessments undertaken by other competent regulatory authorities may be used to avoid duplication. However, genetically modified foods that have already been approved for safety in one country may have to be re-evaluated in a different country to address any significant differences that may arise with respect to consumption patterns, mode of cultivation of crop, form in which the food is consumed and processing methods.


Chassy, B. M. 2002. Food safety assessment of current and future plant biotechnology products. In: Biotechnology and Safety Assessment, Thomas, J. A and Fuchs, R. L. eds., 3rd edition, Academic Press, CA, USA, 95-96.

Codex Alimentarius Commission. Guidance for the conduct of food safety assessment of foods derived from recombinant-DNA plants. CAC/GL 45-2003.

EFSA, 2004. Guidance document of the scientific panel on genetically modified organisms for the risk assessment of genetically modified plants and derived food and feed.

European Food Safety Authority. 2009. Consolidated presentation of the joint Scientific Opinion of the GMO and BIOHAZ Panels on the “Use of Antibiotic Resistance Genes as Marker Genes in Genetically Modified Plants” and the Scientific Opinion of the GMO Panel on “Consequences of the Opinion on the Use of Antibiotic Resistance Genes as Marker Genes in Genetically Modified Plants on Previous EFSA Assessments of Individual GM Plants”. URL

Jonas, D. A., Elmadfa, I., Engel, K. J., Kozianowski, G., König, A., Műller, D., Narbonne, J. F., Wackernagel, W. and Kleiner, J. 2001. Safety considerations of DNA in food. Annals of Nutrition and Metabolism 45, 235-254.

König, A., Cockburn, A., Crevel, R. W. R., Debruyne, E., Grafstroem, R., Hammerling, U., Kimber, I., Knudsen, I., Kuiper, H. A., Peijnenburg, A. A. C. M., Penninks, A. H., Poulsen, M.,Schauzu, M., Wal, J. M. 2004. Assessment of the safety of food derived from genetically modified (GM) crops. Food Chemistry and Toxicology, 42, 1047-1088.

Kuiper, H.A. and Kleter, G.A. 2003. The scientific basis for risk assessment and regulation of genetically modified foods. Trends in Food Science and Technology 14, 277-293.

Kuiper, H.A., Kleter, G.A., Noteborn, H.P.J.M., Kok, E.J., 2001. Assessment of the food safety issues related to genetically modified foods. Plant Journal 27, 503–528.

OECD, 1993. Safety Evaluation of Foods Derived by Modern Biotechnology: Concepts and Principles. Organization for Economic Cooperation and Development (OECD), Paris.

Setchell, K.D.R., Gosselin, S.J., Welsh, M.B., Johnston, J.O., Balistreri, W.F., Krammer, L.W., Dresser, B.L. and Tarr, M.J. 1987. Dietary estrogens – a probable cause of infertility and liver disease in captive cheetahs. Gastroenterology 93, 225-233.

Wang, G., Kuan, S.S., Francis, O.J., Ware, G.M. and Carman, A.S. 1990. A simplified HPLC method for the determination of phytoestrogens in soybean and its processed products. Journal of Agricultural and Food Chemistry 38, 185-190.

World Health Organization. Are GM Foods Safe?