Safety Evaluations of the Cry1Ia1 Protein Found in the Transgenic Potato ‘SpuntaG2’

The transgenic potato ‘SpuntaG2’ (Solanum tuberosum), which is resistant to potato tuber moth (Phthorimaea operculella), was subjected to protein safety evaluations including protein equivalency tests for the Cry1Ia1 protein from ‘SpuntaG2’ and bacterially produced Cry1Ia1, toxicity and allergenicity evaluations of Cry1Ia1 protein, and compositional equivalency of ‘SpuntaG2’ compared with non-transgenic ‘Spunta’. Western blot analysis and biological activity assays showed molecular and functional equivalency between ‘SpuntaG2’-derived Cry1Ia1 protein and bacteria-derived Cry1Ia1 protein. Comparison of the Cry1Ia1 amino acid sequence to known amino acid sequences revealed no significant homology to known toxins or known allergens. Acute toxicity studies using rodents were used to calculate an acceptable daily intake (ADI) value of 20 mg kg body weight per day. The ADI value was then used to calculate a margin of exposure (MOE) of 2,222,222, which is more than 22,000 times greater than the commonly used target MOE of 100. Digestibility and thermostability assays determined that Cry1Ia1 was fully digested within 30 s of exposure to pepsin and inactive after 3 to 4 minutes at 100 8C, indicating that it would not be a potential allergen. Compositional analyses revealed no difference between ‘SpuntaG2’ and non-transgenic ‘Spunta’. These results strongly indicate that the Cry1Ia protein and the transgenic potato ‘SpuntaG2’ is not a human health risk. ‘SpuntaG2’, a transgenic potato resistant to potato tuber moth, has been previously described by Douches et al. (2002). ‘SpuntaG2’ contains two foreign genes: cry1Ia1, a gene derived from the soil-borne bacterium Bacillus thuringiensis and nptII, a gene derived from the bacterium Escherichia coli (Beck et al., 1982). Cry1Ia1 is a protein that is toxic to potato tuber moth, and acts by binding to specific receptors on the membranes of epithelial cells in the target insect’s midgut. Binding to the receptors causes an ion channel to form between the cell cytoplasm and the external environment, resulting in the loss of osmotic regulation and insect death (Gill et al., 1992). Received for publication 19 Jan. 2010. Accepted for publication 15 Apr. 2010. Corresponding author. E-mail: [email protected]. J. AMER. SOC. HORT. SCI. 135(4):325–332. 2010. 325 Another foreign gene that is expressed in ‘SpuntaG2’ is nptII (neomycin phosphotransferase), which is a protein that confers resistance to the antibiotic kanamycin and is used as a selectable marker during the transformation process. The 29-kDa neomycin phosphotransferase enzyme inactivates aminoglycoside antibiotics by transferring the g-phosphate from ATP to the respective antibiotic. A complete characterization of the ‘SpuntaG2’ transgenic event has been conducted (Zarka et al., 2010) and they showed that there is a single copy of each gene with no interruption of endogenous genes during the transformation process. That article is the first one in a three-article series and contains a complete research rationale. We report here, in the second article of the series, on the protein safety evaluations conducted on the foreign proteins in ‘SpuntaG2’ plants. The Cry1Ia1 protein analysis includes protein equivalency tests for the Cry1Ia1 protein from ‘SpuntaG2’ and bacterially produced Cry1Ia1, toxicity and allergenicity evaluations of Cry1Ia1 protein, compositional equivalency of ‘SpuntaG2’ compared with non-transgenic ‘Spunta’, molecular and functional equivalency between ‘SpuntaG2’-derived Cry1Ia1 protein and bacterial-derived Cry1Ia1 protein, and homology analysis to known toxins or known allergens. These experiments were conducted in accordance with the guidelines set forth by Codex Alimentarius (2003), which represent the international scientific standard on the process and data requirements for assessment of food safety relating to genetically modified plants. Particular attention was given to the requirements pertaining to food and feed safety, as contained in sections 5.1 and 5.2 of the guidelines for the use of genetically modified organisms. The other foreign protein in ‘SpuntaG2’ is the NPTII protein and has been previously evaluated and proven safe by the European Food Safety Authority (2007) and Fuchs et al. (1993b). The amount of NPTII expression in ‘SpuntaG2’ was evaluated and compared with other reported levels of expression. Materials and Methods PRODUCTION OF PURIFIED CRY1IA1 PROTEIN. The expression vector system pET28b (Novagen, Madison, WI) was used for cloning and expression of the cry1Ia1 gene. pET28b was digested with the BamHI restriction enzyme and the linearized plasmid was then ligated to a BamHI fragment of the cry1Ia1 gene. The resulting plasmid was sequenced and named pSPUD56. It was then transformed into the E. coli strain BL21(DE3) (Novagen), a general-purpose expression host. The protein was expressed with an N-terminal histidine tag, which enabled it to be purified with Ni-NTA agarose (Qiagen, Valencia, CA). Ten liters of E. coli BL21(DE3) pSPUD56 were produced at the Protein Expression Laboratory of Michigan State University (East Lansing) according to the following protocol: a 250-mL overnight culture of BL21(DE3) pSPUD56 was used to inoculate a 10-L vessel containing HM medium. The HM medium was prepared by dissolving 9 g of KH2PO4, 6 g of K2HPO4, 4 g of Na2HPO4, and 3 g of (NH4)2HPO4 in 940 mL of distilled H2O and autoclaving this for 30 min at 121 C on liquid cycle. After cooling the solution to <40 C, 40 mL of autoclaved 50% glucose, 15 mL of autoclaved 1 M MgSO4, and 5 mL of sterile trace element solution were added. The trace element solution consisted of 10 g of FeSO4, 2.5 g of ZnSO4 7H2O, 1 g of CuSO4 5H2O, 1 g of MnSO4 5H2O, 1 g of CoCl12 6H2O, 1 g of Na2MoO4 2H2O, 5 g of CaCl12 2H2O, and 0.2 g of Na2B4O7 10H2O dissolved in 1 L of 5M HCl. The 10-L vessel containing the HM medium and BL21(DE3)pSPUD56 culture was agitated at 500 rpm with an air flow rate of 1 L min while the pH was controlled by adding ammonia. After 12 h of culture, IPTG was added to a final concentration of 0.5 mM to induce production of the Cry1Ia1 protein. The dissolved oxygen was maintained below 10% (w/v) by adding 2.8 M glucose as needed. After 6 h of induction, the cells were harvested by centrifugation and the wet pellet was stored at –80 C until purification. The E. coli pellet was taken from –80 C and resuspended in PN buffer (50 mM NaH2PO4 and 0.3 M NaCl, pH 7.8) containing 1% Triton X-100 (Bio-Rad, Hercules, CA) and 1 mg mL lysozyme in a centrifuge tube. The cell suspension was prepared for and purified with Ni-NTA agarose (Qiagen) according to the manufacturer’s instructions. The protein was eluted by adding PN buffer containing 0.5% Triton X-100 and 0.3 M imidazole. The eluate (containing the purified protein) was collected. The protein solution was then concentrated and simultaneously changed to PEN buffer [50 mM NaH2PO4, 0.3 M NaCl, 0.5 mM EDTA, pH 7.8, Protease Inhibitor (Roche Applied Science, Indianapolis), and 10% glycerol], using centrifugational filter devices (Amicon Ultra-15; Millipore, Bedford, MA). The concentration of protein samples was determined by using a commercially available Bradford Reagent (Bio-Rad) according to manufacturer’s instructions. Purified protein preparations were stored at –80 C. The purified protein preparation from the previous step was thawed and concentrated further to 160 mg mL. A sample was diluted and separated by electrophoresis on a 10% Tris Glycine SDS-PAGE (Bio-Rad) according to Laemmli (1970). After electrophoresis, the gel was stained with SimplyBlue SafeStain (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions to confirm the size and purity of the sample. Based on the intensity of the Cry1Ia1 band, the preparation was judged to be >90% pure. EQUIVALENCY OF IN VITRO CRY1IA1 PROTEIN FROM E. COLI AND CRY1IA1 PROTEIN FROM ‘SPUNTAG2’. Non-transgenic ‘Spunta’ and ‘SpuntaG2’ proteins were compared by western blot analysis with E. coli-derived Cry1Ia1 protein previously described by Li et al. (1999). The biological activities of Cry1Ia1 protein purified from E. coli and Cry1Ia1 protein contained in sap from ‘SpuntaG2’ were also compared. While the target insect for ‘SpuntaG2’ is the potato tuber moth, there currently is no known artificial diet capable of maintaining this insect during the larval stages. However, Cry1Ia1 is also toxic to the tobacco hornworm (Manduca sexta), an insect for which eggs and an artificial diet have been developed and are commercially available. Therefore, the tobacco hornworm was used for this study. The eggs were purchased from the North Carolina State University Entomology Insectary (Raleigh). Once the first instar larvae emerged, the insects were placed in 24-well plates (Cellbind; Corning, Corning, NY) containing 1 mL/well commercially prepared tobacco hornworm diet (Carolina Biological Supply, Burlington, NC) and one newly hatched hornworm per well. Purified E. coli Cry1Ia1 was mixed into the wells at concentrations of 0.1, 0.5, 1.0, 1.5, and 2.5 mg mL. As a control, 100 mL of ddH2O was mixed with the diet (0 mg of Cry1Ia1 protein). Twenty-four replicate wells were prepared for each protein concentration. The larvae were added and the plates were placed at room temperature for 72 h. The surviving larvae were recorded. 326 J. AMER. SOC. HORT. SCI. 135(4):325–332. 2010. To determine the activity of the Cry1Ia1 protein in ‘SpuntaG2’, leaf tissue was extracted from the non-transgenic ‘Spunta’ control or from ‘SpuntaG2’. Using a mortar and pestle, leaf tissue samples were ground and the sap was removed by pipetting. This sap sample was centrifuged at 2000 gn for 5 s to pellet any large debris. The sap was transferred to a new tube. The sap was weighed and it was determined that the sap represented 0.001 g of fresh wet leaf tissue weight per microliter. The liquid sap was then mixed into the commercial diet in the wells (10 mL = 0.01 g tissue, 25mL = 0.025 g tissue, 50 mL= 0.05 g tissue, and 100 mL = 0.1 g tissue). The larvae were added and the plates were placed at room temperature for 72 h. The surviving larvae were then recorded. TOXICOLOGY. Homology of Cry1Ia1 protein to known toxins was assessed using the approach described by Rice et al. (2008). The amino acid sequence of the Cry1Ia1 protein was used to query the database using the algorithm, Basic Local Alignment Search Tool Protein (BLASTP) (Altschul et al., 1997). The BLASTP algorithm was implemented via the National Center for Biotechnology Information (NCBI, 2010). The default parameters were used. This implementation of the BLASTP algorithm searches the protein-nr public database of protein sequences. The delimiting phrase ‘‘toxin OR toxic NOT insecticidal NOT pesticidal NOT delta endotoxin’’ was used. This delimiter was used because an initial search using just the phrase ‘‘toxin OR toxic’’ returned only Cry proteins, and was therefore not informative. The acute toxicity of Cry1Ia1 protein was evaluated by MPI Research, Mattawan, MI (MPI Research, Inc., 2008). Purified Cry1Ia1 protein was used and the study was conducted in compliance with the U.S. Food and Drug Administration (USFDA, 2008). USFDA and international guidelines were followed (International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use, 1997; USFDA, 2004). One treatment group of 10 male and 10 female Crl:CD1 (Icr) mice was administered the test article, Cry1IA1 protein, at a dose level of 2000 mg kg and at a dose volume of 12.5 mL kg. One additional group of 20 animals (10 males and 10 females) served as the control and received the vehicle, control buffer, at a dose volume of 12.5 mL kg. The test article or vehicle was administered via oral gavage, once on Day 1. Observations for mortality, morbidity, injury, and the availability of food and water were conducted twice daily for all animals. Clinical observations were conducted once on day 1 ( 1 h post dose), day 7, and day 14. Body weights and food consumption were measured and recorded weekly. At study termination, necropsy examinations were performed, organ weights were recorded, and selected tissues were preserved. ALLERGENICITY. The search for sequence similarity between the Cry1Ia1 protein and known allergens was conducted by comparing the amino acid sequence of Cry1Ia1 with the database of known allergens, AllergenOnline (University of Nebraska-Lincoln, 2010). Sequence similarity was done using the FASTA3 algorithm (Pearson, 2000), accessed through the AllergenOnline website. The website conducts two types of sequence searches: 1) an overall FASTA search of the AllergenOnline database, and 2) an 80-amino acid sliding window search in which sequential 80-amino acid segments of the protein are generated and each is used to query the AllergenOnline database. In these sequence searches, Codex Alimentarius (2003) considers allergenicity to be a possibility if there is 35% (or more) identity between the protein in question and a known allergen in a segment of at least 80 amino acids. In addition to sequence analyses, Codex Alimentarius (2003) also recommends a test to determine the ability of pepsin to digest the protein in question. Therefore, the digestibility of the Cry1Ia1 protein was investigated in an in vitro digestibility assay according to the procedure followed by Fu et al. (2002). To initiate a time course experiment, 665 mL of SGF solution was incubated at 37 C for 2 min, and then 35 mL of the purified Cry1Ia1 protein at 5 mg mL was added; 100 mL of sample was removed and added to a separate tube containing 35 mL of 200 mM NaHCO3 (pH 11.0) and 50 mL 3· Laemmli sample buffer. The neutralized samples were analyzed by SDS polyacrylamide resolving gel run according to Laemmli (1970) along with a 0 time point, which consisted of 0.7 mg of Cry1Ia1 protein that was added to already neutralized SGF. Results were visualized by staining the gel with Coomassie Blue. Another criterion for assessing the allergenic potential of a protein is its thermostability as measured by activity (Delaney et al., 2008). The stability of the Cry1Ia1 protein after being exposed to heat was studied using tobacco hornworm as a test insect due to the availability of an artificial diet capable of maintaining this insect during the larval stages. The tobacco hornworm eggs were purchased from the North Carolina State University Entomology Insectary (Raleigh). Cellbind 24-well plates were used for this study with two wells per dose of Cry1Ia1 per time point, and only one well per time point, up to 3 min, was used for the control. Purified Cry1Ia1 protein was placed in a hot block set at 100 C. At 0, 1, 2, 3, 4, and 5 min a sample (10 or 100 mg mL in 100 mL of ddH2O) was removed and mixed into the diet (Carolina Biological Supply). As a control, 100 mL ddH2O was mixed with the diet (0 mg of Cry1Ia1 protein). Two newly hatched first instar larvae were added to each well and the plates were placed at room temperature for 72 h. The surviving larvae were then recorded. COMPOSITIONAL ANALYSIS OF ‘SPUNTAG2’. The nutritional composition of ‘SpuntaG2’ was analyzed in comparison with non-transgenic ‘Spunta’. This analysis, in conjunction with a comparison of agronomic characteristics, are the primary analyses recommended by Codex Alimentarius (2003) for the assessment of unintended effects resulting from the transformation process. The components analyzed were based on those specified by the Organization for Economic Co-operation and Development (OECD, 2002; Rogan et al., 2000). Potatoes were sampled in South Africa from four replicate plots of ‘SpuntaG2’ and ‘Spunta’ at two locations (Kokstad and Petrus Steyn) and from three replicate plots at Roodeplaat. Freshly harvested tubers from each plot were placed in individual piles. ‘Spunta’ and ‘SpuntaG2’ tubers ( 3 kg) were randomly removed from each pile and placed in bags. About 1 kg of fresh tubers was removed from the bags and the tubers were washed and diced into cubes about 1 cm in size. The diced tuber pieces were placed in plastic bags, frozen at –20 C, and sent to and analyzed by Agricultural Research Council-Irene Analytical Services (Pretoria, South Africa), a South African National Accreditation System accredited laboratory. The data were analyzed by two-way analysis of variance, with Buonferroni post-tests, using GraphPad PRISM software (version 5.0 for Mac; GraphPad PRISM Software, San Diego). Nutrients that were not detected were not included in the statistical analysis. J. AMER. SOC. HORT. SCI. 135(4):325–332. 2010. 327 A rodent feeding study entitled, ‘‘Testing the nutritional adequacy of a genetically modified ‘Spunta’ potato using the rat model,’’ was conducted by the Agricultural Research Council Toxicology (Pretoria, South Africa) as an additional assurance of nutritional adequacy and absence of unintended effects. Before the study described in this article, a pilot study was conducted to determine the nutritional adequacy of various test doses of nontransgenic potato on rats. This study showed no adverse nutritional effect of 30% (w/v) potato, the highest dose tested, and therefore this dose was chosen for the main study. Pellets containing 30% (w/v) potato meal were made by mixing the freeze-dried ‘Spunta’ or ‘SpuntaG2’ potato tubers with the milled standard rodent pellets (Onderstepoort Biological Products, Pretoria, South Africa) and reformulating pellets from the mixture. Standard proximate analysis of the freeze-dried tubers as well as test and control pellets was done to assure nutritional equivalence. Young Sprague-Dawley rats (female n = 30, male n = 30), with masses between 49 and 78 g, were obtained from National Health Laboratory Services (Johannesburg, South Africa) and kept in an environmentally controlled small animal facility. During the adaptation period of 1 week, tap water and standard rat pellets were provided ad libitum. The rats were divided randomly into six groups (n = 10) according to the following scheme: control female and male groups: standard pellets only; ‘Spunta’ female and male groups: pellets containing 30% (w/v) ‘Spunta’; ‘SpuntaG2’ female and male groups: pellets containing 30% (w/v) ‘SpuntaG2’. During the main trial (90 d) rats were kept in single cages facing one another and tap water and feed were provided ad libitum. The individual body masses as well as the feed intake were recorded twice weekly to compute and compare some of the performance parameters (body mass gain, food intake, final body mass, and feed efficiency). Temperature was maintained at 23 C to 24 C, relative humidity varied between 43% and 65%, and lighting was controlled on a 12/12-h light/dark cycle (lights on at 0600 HR; lights off at 1800 HR). The experiment was ended by euthanizing the rats (overdose of pentobarbitone sodium 200 mg mL at a dose of 1–2 mL kg intraperitoneally), immediately followed by collecting blood from the heart (1.5 mL with a 24 G needle) and removing the liver, heart, kidney, spleen, lungs, and ovaries or testes for weighing. The blood was centrifuged (at 1200 gn for 10 min) and the serum was collected and submitted for clinical chemistry and hematology on whole blood. D E T E R M I N A T I O N O F N P T I I EXPRESSION IN ‘SPUNTAG2’. A commercial NPTII ELISA kit (catalog no. PSP 73000; Agdia, Elkhart, IN) was used to determine the level of NPTII protein present in ‘SpuntaG2’. Freeze-dried leaves were ground in sample buffer (PEB1, supplied by the manufacturer) at a concentration of 10 mg mL and were then diluted to 2.5 mg mL dry weight. This dilution was determined in previous experiments to provide readings in the linear range of response for the assay. Samples were loaded onto pre-coated ELISA plates (100 mL/well) along with standards comprised of 2-fold serial dilutions of positive controls, as well as negative control samples on each plate. The optical density of each well was read with a microplate reader (model 50; Varian Cary, Palo Alto, CA). Four replicate experiments, done on different dates, were conducted. Concentrations of NPTII protein were based upon standards run on the same plate, at the same time.