Mason Holtel
November 1, 2012
Genetics 303
Professor Bert Ely
Term Paper- Allergenicity of Genetically Modified Crops
Genetically modified (GM) crops are agricultural crops that have been genetically altered to enhance or introduce a specific trait that will not occur naturally. The most prevalent reason for altering a crop’s genetic material within the past two decades has been to benefit society with a more efficient food source. Most often GM crops are produced by inserting a gene from one organism into the genome of the new organism. This transgene now has the capacity to interact with its new organism in the same way that it did with the old organism. It may aid to combat disease more efficiently, or it may produce proteins that inhibit consumption by insects or other pests. Many GM crops even contain enhanced proteins that provide more nutrition to its consumer than the naturally occurring crop.
Regardless of its advantage, GM crops have become more widespread over the last decade or so. According to James (2011), GM crops have been used for the past sixteen years and are currently being used in twenty-nine countries, spanning over 160 million hectares. The globe has had a growing need for GM crops in order to produce higher yields of more nutritious foods to feed the growing human population (Young et al. 2012). However, skeptics along the way have been wary of altering the genetics of consumable products. They fear that these alterations could have unpredictable effects on the consumers and even the environment.
One of the greatest fears regarding the introduction of genetically modified foods into the diet of a population is the concern that these new proteins will cause cross-reactivity of allergens. Cross-reactivity is where allergic reactions are triggered, not from the allergen, but from a similar protein that can bind to the corresponding antibody. For example, many doctors tell patients to avoid all types of shell-fish because the antigens are all so similar they can activate the same allergic reaction. In humans, Immunoglobulin E (IgE) is the class of antibody that defends against various allergens (Mishra et al. 2012; Nakajima et al. 2010). If a protein binds to IgE, there is a chance that an allergic reaction could occur.
To prevent occurrences of cross-reactivity in GM crops, various organizations such as the Food and Agriculture Organization (FAO), World Health Organization (WHO), and Codex Alimentarius Commission (Codex) have defined guidelines to determine whether a new GM crop can be commercialized(Young et al. 2012; Mishra et al. 2012). This test is intended to prevent cross-reactivity and further harm to both human and animal consumers, as well as the local environment, by comparing the genetic material of GM crops with a database of known allergens. Codex guidelines define a GM crop as “safe for commercialization” if there are no polypeptide chains that share 35% identity with a known allergen (Young et al. 2012; Mishra et al. 2012; Nakajima et al. 2010). This standard means that if a particular amino acid sequence over 80aa aligns with a known allergen by 35% or more, that sequence is considered a “potential allergen,” which would call for more studies. Furthermore, Codex has broadened the scope of amino acid sequences by including all six DNA reading frames between stop codons, not just regions that code for proteins (Young et al. 2012). This precaution is established to prevent any dormant, uncoded allergens from appearing in the near future by methods of mutation.
The most common method of testing for allergenicity of new GM crops is running its amino acid sequence with databases of known allergens in silico, or on the computer. Multiple allergen databases are used to confirm thoroughness and accuracy. If a GM crop passes the test of less than 35% identity, it is considered safe for commercial use. If sequences do contain 35% or more identity with known allergens, further in vitro testing is conducted to confirm or deny allergenicity (Young et al. 2012; Nakajima et al. 2010).
One particular study conducted by Young et al. (2012) challenged the efficiency of this established in silico test. These scientists compared the genetic sequences of non-transgenic (unaltered genome) soybean (G. max), maize (Z. mays), and two subspecies of rice (O.s. ssp. indica and O.s ssp. japonica) with the AllergenOnline Database 11. The human genome was also used as a non-plant control. The AllergenOnline Database 11 (FARRP 11) contains 1491 sequenced allergens monitored by the Food Allergy Research and Resource Program (FARRP). Under the stipulations of Codex, all portions of DNA between stop codons were sequenced, not just regions that code for proteins. The purpose of this study was to determine whether guidelines to govern potential allergenicity are too conservative, meaning that too many potential allergens are marked during the in silico study. The hypothesis stated that even when common non-transgenic crops are sequenced and compared with allergen databases, far too many segments are flagged as potential allergens (Young et al. 2012).
Young et al. (2012) began their study by collecting genomes of all the organisms used. Z. mays, G. max, and O.s. ssp. japonica were collected from the Phytozome website, O.s ssp. indica came from the Beijing Genomics Institute, and the human genome was collected from the National Center for Biotechnology Information (NCBI) (Schnable et al. 2009; Yu et al. 2002). Sequences of amino acids were obtained by scanning all portions of DNA between stop codons and translating each triplet into an amino acid using a standard genetic codon table. Once amino acid sequences were collected, they were run through an algorithm machine, FASTA, to be compared with the known allergen list from FARRP 11. Gaps in the sequences were considered mismatches (Young et al. 2012).
An alternative study conducted by Mishra et al. (2012), ran the same experiment, only they tested specific transgenes rather than non-transgenic crops. These scientists tested allergenicity of osmotin, glucanase, and chitinase which are all transgenes to aid plants in defense against fungal infections. They also ran the sequences of superoxide dismutase and glycine betaine aldehyde dehydrogenase which are transgenes that relieve oxidative and osmotic stress (Mishra et al. 2012). Both experiments conducted the experiment in the same way, only differing in the substances being tested for allergenicity.
Results obtained from both experiments showed an immense amount of possible allergens. Young et al. (2012) yielded results showing thousands of possible allergens for each species and subspecies tested including the human genome (Table 1). Since Codex guidelines set the standard for allergens at 80aa or greater, only peptide chains pertaining to this stipulation were included in the analysis. Also, naturally occurring allergens within the plant species were screened, meaning that already known allergens of the non-transgenic substance were omitted in sequencing. Z. mays contained the greatest number of potential allergens at 33,475. G. max showed 3,393 cases of potential allergenicity. Both strains of rice, O.s. ssp. indica and japonica, were each also marked with over ten thousand matches to the FARRP 11 database. Surprisingly, the human genome was marked 21,968 times for potential allergenicity (Young et al. 2012).
Summary of FASTA analysis against FARRP 11 Database
Genome / Polypeptides chains ≥80aa / ≥35% identityZ. mays / 6,993,325 / 33,475
G.max / 1,434,209 / 3,393
O.s. ssp. indica / 1,196,644 / 10,034
O.s. ssp. japonica / 1.064,957 / 11,479
H. sapiens / 5,882480 / 21,968
The following table was produced using data from the experiment of (Young et al. 2012). Table 1 shows the number of polypeptide chains within the genome of certain species of crops. The polypeptide chains are hypothetical in the sense that all DNA regions are considered possibilities for allergens, not just the protein coding regions. Thirty-five percent or more identity means that this number of polypeptide chains over 80aa matched a known allergen from the FARRP 11 database.
Likewise, Mishra et al. (2012) yielded results stating that the transgenes tested matched with several known allergens. The manganese superoxide dismutase gene showed sequence identity of 85.09% with the known allergen Hevea brasiliensis. The rice chitinase gene showed a 68.57% identity with allergens found in rubber and chestnut plants. The glycine betaine aldehyde dehydrogenase gene showed a 43.70% with various known allergens. Osmotin showed the greatest identity with allergens reaching 91.87% identity match with a known allergen. Finally, alfalfa beta-1, 3-glucanase and wheat beta-1, 3-glucanase genes showed 38.96% and 36.33% identity with known allergens respectively (Table 2).
Sequence Identity between a Transgene and a Known Allergen
Transgene / Identity Match with Known AllergenManganese superoxide dismutase / 85.09%
Rice chitinase / 68.57%
Glycine betaine aldehyde dehydrogenase / 43.70%
Osmotin / 91.87%
Alfalfa beta-1,3-glucanase / 38.96%
Wheat beta-1,3-glucanase / 36.33%
The following data was produced using data found from Mishra et al. (2012). Table 2 shows the percent identity match between a certain transgene and a known allergen. Based on these identities, none of the transgenes would be considered for commercial use as further testing must be used to test for allergenicity.
This procedure, as explained by both experiments, does contain complications. According to Young et al. (2012), four non-transgenic, everyday crops eaten by millions of people for thousands of years would fail the Codex safety assignment. These scientists state that there are far too many false positives for allergenicity. They explain that running further tests for each and every potential allergen would be extremely inefficient. Likewise, although Mishra et al. (2012) found all of their transgenes to be potential allergens, the in silico tests were highly inaccurate. For example, further studies of in vitro testing of an IgE binding assay proved that osmotin was indeed cross-reactive with apple and tomato allergens, confirming the results found from the in silico study (Sharma et al. 1999). However, other studies proved that rice chitinase is not accurately reflected as a cross-reactive allergen through in silico sequence testing (Alenius et al. 1995; Blanco et al. 1999). Therefore, both studies call for a more accurate means of studying cross-reactivity.
Nakajima et al. (2010) demonstrates the methods in which allergenicity tests are more accurately performed. In this particular experiment the transgene Cry3Bb1 was extracted from the GM corn crop MON863 and tested for allergenicity by means of an IgE binding test. This test was conducted by extracting the DNA containing Cry3Bb1 from MON 863 and replicating using polymerase chain reaction (PCR). Assay plates were prepared by collecting serum samples of thirteen corn allergy patients from the United States, fifty-five corn allergy patients from Japan, and three control patients from Japan. Each serum sample was inserted into an E. coli culture for the IgE binding assay. Each sample was then coated with a purified strain of Cry3Bb1. Nakajima et al. (2010) then conducted a Western blot, to prove that IgE antibodies did not bind to the Cry3Bb1 gene, meaning that Cry3Bb1 was not recognized as an allergen. The scientists concluded that Cry3Bb1 was not a threat to allergic reactions by any patient in the study. This determined that MON863 was just as safe as any non-transgenic corn crop in this particular locus. Had binding occurred, they stated, further tests such as skin tests and oral challenges would have been conducted to confirm the IgE binding test (Nakajima et al. 2010).
As shown in all three of these experiments, safety requirements for the commercialization of GM crops fall under strict scrutiny. Young et al. (2012) and Mishra et al. (2012) both call for a less conservative method of testing. After running an in silico test, there are far too many false positives, each of which requires further testing to confirm. These tests can be time consuming as each and every locus in question must be run through an extensive series of IgE binding tests (Nakajima et al. 2010). With thousands of potential allergens even in non-transgenic crops, complete and thorough testing for GM crops could take years of scrutiny to commercialize. One idea to better the testing requirements is to forget the stipulation that requires all hypothetical polypeptides from being tested (Young et al. 2012). These scientists claim that this step is unnecessary since sequences derived from stop- to-stop codons are highly unlikely to produce an allergen. Furthermore, there is no record of allergens becoming active from stop-to-stop codons in the history of any crop species. Finally, Mishra et al. (2012) suggests that the future in 3D analysis can aid in predicting allergenicity. With the help of improving computer programs, new polypeptide sequences can be modeled and displayed to determine if the active site of the new protein sequence will fit into an IgE binding site similar to its cross-reactive allergen.
Testing for allergenicity by means of in silico studies proves to be a good starting point. However, due to its conservative approach and high rates of false positives, vast numbers of in vitro tests must be run in order to accurately prove the allergenicity of a new polypeptide sequence. Therefore, many scientists request stricter guidelines from the Codex safety assessment of allergenicity in order to obtain more efficient predictions of future allergens in GM crops.
Literature Cited