Safety Assessment of Bollgard II Cotton Event 15985

Transcription

1 Executive Summary Safety Assessment of Bollgard II Cotton Event Bollgard II cotton event was developed by Monsanto Company to produce the Cry2Ab2 insect control protein, which provides effective season-long control of key lepidopteran insect pests. This product was produced by re-transformation of Bollgard cotton event 531, which produces the Cry1Ac insect-control protein and the NPTII selectable marker protein. Therefore, Bollgard II cotton produces two proteins for effective control of the major lepidopteran insect pests of cotton, including the cotton bollworm, tobacco budworm, pink bollworm, and armyworm. Bollgard II cotton also produces the β-d-glucuronidase (GUS) marker protein. In addition, Bollgard II cotton provides a more effective insect resistance management program compared to single gene products. Bollgard cotton has been grown globally on more than 32 million acres since commercial introduction in 1996 (James, 2002). The primary benefits that have resulted from the use of Bollgard cotton are reduced insecticide use, improved control of target insect pests, improved yield, reduced production costs, and improved profitability for cotton growers (Edge et al., 2001; Carpenter and Gianessi, 2001; Betz et al., 2000; Economic Research Service/USDA, 2000; Falck-Zepeda et al., 1998; Falck-Zepeda et al., 2000; Fernandez- Cornjeo and McBride, 2000; Klotz-Ingram et al., 1999; Traxler and Falck-Zepeda, 1999; Xia et al., 1999). With the addition of a second insect protection protein, Bollgard II cotton provides increased control of cotton bollworm, as well as certain secondary insect pests of cotton, including armyworm (U.S. EPA, 2002). Furthermore, along with the other components of Monsanto s insect resistance management program, combining the Cry2Ab2 and Cry1Ac proteins in a single product provides an additional tool to delay the development of insect resistance to Cry proteins in cotton. The Cry2Ab2 protein produced in Bollgard II cotton event is derived from the naturally occurring soil bacterium Bacillus thuringiensis (B.t.). Microbial formulations of Bacillus thuringiensis, which include the Cry2A class of proteins, have been registered in numerous countries worldwide and have been safely used for control of lepidopteran insect pests for more than 40 years (Lüthy et al., 1982; Baum et al., 1999; IPCS, 1999; Betz et al., 2000). B.t. microbial formulations have been shown to be specific to the target insect pests and do not have deleterious effects to non-target organisms such as beneficial insects, birds, fish, and mammals, including humans (U.S. EPA, 1988; U.S. EPA, 1998). Therefore, there is a history of safe dietary and occupational exposure to Cry proteins derived from B.t., including those of the Cry2A class. The GUS protein present in Bollgard II cotton was used as a marker to facilitate the selection of Cry2Ab2-producing plants. The GUS protein served no other purpose and has Bollgard and Bollgard II are registered trademarks of Monsanto Technology LLC. August

2 no known insect control properties. The history of safe use of the GUS protein is extensive. Human exposure to the GUS protein is commonplace through intestinal epithelial cells and intestinal microflora, bacterial exposure, and numerous foods containing the GUS protein with no known harmful effects (Gilissen et al., 1998). The Cry2Ab2 and GUS proteins in Bollgard II cotton event are present at very low levels in cottonseed and are expected to be absent or inactivated in highly processed cotton food and feed products. The safety of the introduced proteins has been assessed through the confirmation of a history of safe food and feed use of the proteins or highly similar proteins, the determination of no significant allergenic potential of the introduced proteins, and the determination of no significant toxic potential of the introduced proteins. Furthermore, there will be no significant consumption of these proteins in foods derived from Bollgard II cotton due to the extensive processing and refinement of cottonseed oil and cotton-derived food products. An assessment of the nutritional and compositional equivalence of Bollgard II cotton to conventional cotton varieties was performed on 48 components of cottonseed, oil, and meal. These analyses included protein, fat, moisture, calories, minerals, amino acids, cyclopropenoid fatty acid, and gossypol levels. Results of these extensive compositional analyses demonstrated that the levels of the important nutritional and anti-nutritional components in Bollgard II cotton event are comparable to the parental variety and are within established ranges for commercial cotton varieties. It is concluded that Bollgard II cotton event is not materially different in composition, safety, or any relevant parameter from cotton now grown, marketed, and consumed. The following summary provides information on the methods used to develop Bollgard II cotton event and a summary of the food, feed, and environmental safety studies performed. On the basis of these evaluations, Bollgard II cotton and its processed fractions were found to be substantially equivalent to conventionally bred cotton, taking into consideration the natural variation observed among cotton varieties, with the exception of the production of the Cry1Ac, Cry2Ab2, NPTII, and GUS proteins. Previous studies established the food, feed, and environmental safety of the Cry1Ac and NPTII proteins produced in Bollgard cotton, and more recent studies have confirmed that the Cry2Ab2 and GUS proteins produced in Bollgard II cotton are also safe for human and animal consumption and to the environment. August

3 Introduction Cotton is the leading plant fiber crop produced in the world and the most important in the United States. Cotton production in the United States is located primarily in the tier of 15 southern states stretching from California to North Carolina, with approximately 13 million acres grown annually (James, 2002). Lepidopteran insects are the main insect pest problem in cotton. The primary lepidopteran pests infesting cotton are cotton bollworm, tobacco budworm, and pink bollworm. The average percent yield loss due to bollworm and budworm infestation between 1985 and 1995 was 3.7 % (Gianessi and Carpenter, 1999). During the growing season other insects (e.g., cotton boll weevil, lygus bugs, fleahoppers, spider mites, thrips, and aphids) are also present. Bollgard cotton, which produces the Cry1Ac insect control protein, has been adopted broadly by growers since its commercial introduction in 1996, as it provides effective protection from feeding damage by lepidopteran insect pests such as tobacco budworm, pink bollworm, and cotton bollworm (Carpenter and Gianessi, 2001). Bollgard cotton growers typically apply significantly less insecticide to control these pests, realize higher yields, and achieve greater profitability compared to growers using conventional cotton varieties (Fernandez-Cornejo and McBride, 2000). Bollgard cotton has been grown on more than 32 million acres globally since it was introduced in the United States in 1996 (James, 2002). The food, feed, and environmental safety of Bollgard cotton has been reviewed (Hamilton et al., 2002; Monsanto, 2002). The introduction of Bollgard II cotton, producing both the Cry1Ac and Cry2Ab2 proteins, is expected to expand the range of benefits to both growers and the environment. Bollgard II cotton provides equivalent or increased control of the major insect pests of cotton (tobacco budworm, pink bollworm, and cotton bollworm) compared to Bollgard cotton, with additional control of secondary lepidopteran insect pests such as beet and fall armyworm. Combining the Cry2Ab protein with the Cry1Ac protein in Bollgard cotton will also provide an additional tool to delay the development of resistance since these two protein classes have different modes of action (Crickmore et al., 1998). In general, if the second insecticidal protein is sufficiently different in its mechanism of action from the first, and is itself highly efficacious against the target pest species, then insects would need to develop two distinct modes of resistance to survive both proteins, which is highly unlikely. Therefore, Bollgard II cotton, containing both the Cry1Ac and Cry2Ab proteins, provides added protection against the risk of resistance developing in the primary target insect species and is expected to extend the effectiveness of this technology for the grower and prolong the overall benefits already documented for Bollgard cotton. In conclusion, lepidopteran insect pests -- cotton bollworm, tobacco budworm, and pink bollworm -- are the main insect pest problem in cotton production. Bollgard cotton, producing the insecticidal protein Cry1Ac, has been widely adopted by growers because of its efficacy against these pests and demonstrated environmental and economic benefits. The introduction of Bollgard II cotton, producing both the Cry1Ac and Cry2Ab2 proteins, will expand the range of benefits. Furthermore, Bollgard II cotton, in combination with August

4 other components of the insect resistance management program, is expected to significantly delay the development of insect resistance. The following sections describe the molecular characterization of the inserted DNA, the levels of the Cry2Ab2 and GUS proteins, the safety assessment of the Cry2Ab2 and GUS proteins, the compositional analyses of cottonseed, cottonseed oil and cottonseed meal alone and for Bollgard II cotton compared to other cotton varieties and the environmental risk assessment of Bollgard II cotton. Molecular Characterization of Bollgard II Cotton Bollgard II cotton event was generated by re-transformation of cotton meristems of Bollgard cotton event 531, variety DP50B. A particle acceleration plant transformation procedure was used to insert the cry2ab2 insect control coding sequence and the uida marker coding sequence into the Bollgard cotton genome. The purified plasmid vector, PV-GHBK11, is a 8.7Kb high copy number based plasmid containing well-characterized DNA elements for selection and replication of the plasmid in bacteria (Figure 1). The purified, linear DNA was inserted into the Bollgard cotton genome. The linear plasmid fragment only contains two plant gene expression cassettes, each using separate controlling DNA elements essential for production in the cotton plant cells and does not contain the nptii selectable marker gene or origin of replication. The first cassette contains a copy of the cry2ab2 gene encoding the B.t. insecticidal protein Cry2Ab2 and the second cassette contains the uida gene encoding the β-d-glucuronidase (GUS) marker protein to facilitate selection of Cry2Ab2-producing plants. The GUS protein serves no other purpose and has no known insect control properties. The cry2ab2 and uida genes are both under the regulation of the enhanced cauliflower mosaic virus 35S promoter (e35s) (Odell et al., 1985) and the 3 untranslated region of the nopaline synthase gene (NOS 3 ) from Agrobacterium tumifaciens, which provides the signal for mrna polyadenylation. The e35s promoter driving the cry2ab2 gene is also fused to the 5 untranslated leader sequence from the petunia heat shock protein 70 (HSP70) and the chloroplast transit peptide from the Arabidopsis thaliana 5-enolpyruvyl shikimate-3-phosphate synthase gene (CPT2), which is used to direct the protein to the chloroplasts. Integration of DNA into cotton germlings was detected by histochemical staining for GUS protein activity in vascular tissue. Non-transformed tissue was removed and growth of meristems containing the introduced DNA was promoted. The resulting seed from these plants was screened for the production of Cry2Ab2 protein. The molecular characterization of Bollgard II cotton demonstrated that there is one DNA cry2ab2 insert. The single DNA insert in Bollgard II cotton event contains one copy of the cry2ab and uida gene cassettes from the linear DNA PV-GHBK11 used for transformation containing: August

5 the complete cry2ab coding region and cassette, although the restriction site following the NOS 3 polyadenylation sequence in the cassette is not present; and the complete uida coding region and cassette, except that 260 bp of the 5 end of the enhanced CaMV 35S promoter is not present; however, the truncated promoter is functional as demonstrated by production of the GUS protein. Sequencing of the DNA inserted into Bollgard II cotton confirmed the molecular details above. PCR and DNA sequencing verified the 5 and 3 ends of the insert and confirmed that the DNA flanking the insert was native to cotton. Production of the full-length Cry2Ab2 and GUS proteins was confirmed by western blot analysis. Inheritance analysis of the cry2ab2 insert conforms to the expected Mendelian segregation pattern for a single genetic locus. The stability of the insert was demonstrated by Southern blot over four generations of selfing and two generations of backcrossing. In addition, progeny of Bollgard II cotton event have been field tested at multiple sites in the U.S. since No instability of the DNA cry2ab2 insert has been detected during extensive field-testing and commercial seed production of Bollgard II cotton based on the following results: analyses of seed obtained from multi-site trials over four years showed similar levels of the Cry2Ab2 and GUS proteins; the production of the Cry2Ab2 protein has been confirmed by immuno-detection and/or efficacy data under various environmental conditions and in numerous Bollgard II cotton varieties; the insecticidal efficacy has been maintained during the development of this product in the U.S. and other world areas where this product will be commercialized; and the production of the Cry2Ab2 protein has been maintained after transfer of the cry2ab2 gene into different varieties of cotton. These data confirm that the Bollgard II cotton insert is stably integrated in the cotton genome. Cry2Ab2 and GUS Protein Levels in Bollgard II Cotton Plants Enzyme-linked immunosorbent assays (ELISA) (Harlow and Lane, 1988) were developed and optimized to estimate the Cry2Ab2 and GUS protein levels in cottonseed and cotton leaf matrices. Cry2Ab2 and GUS proteins were detected in various plant tissues of Bollgard II cotton plants during the 1998 growing season across eight locations representative of major cotton production regions (Table 1). The Cry2Ab2 and GUS proteins were detected in Bollgard II cotton plants but, as expected, neither protein was detected in the parental control, Bollgard cotton, or in the non-transgenic control. The Cry2Ab2 protein levels estimated in Bollgard II cotton leaf and seed were 23.9 and 43.2 µg/g fresh weight, respectively. The mean protein levels for GUS were 106 and 58.8 µg fresh weight in leaf and cottonseed, respectively. These protein levels are low in August

6 comparison to total protein levels. Levels of the Cry2Ab2 protein were also measured in whole plants collected at the end of the season, and in pollen. In field tests from the 1998 season, mature Bollgard II cotton plants contained an estimated 8.8 µg Cry2Ab2 protein/g fresh weight. The Cry2Ab2 protein was not detected in pollen collected from Bollgard II cotton plants above the limit of detection of the assay (0.25 µg/g fresh weight). Safety Assessment of Cry2Ab2 and GUS Proteins in Bollgard II Cotton Safety assessments of the Cry2Ab2 and GUS proteins produced in Bollgard II cotton event included protein characterization, demonstration of the lack of similarity to known allergens and toxins, the long history of safe consumption of similar proteins, in vitro digestibility, and the lack of acute oral toxicity in mice. Cry2Ab2 is a protein derived from Bacillus thuringiensis and has also been designated Cry2Ab2, CryIIB, CryB2 or CryIIAb (Liang and Dean, 1994; Widner and Whiteley, 1990; Crickmore et al., 1998). In the current nomenclature scheme, Cry protein names are assigned according to amino acid similarity to establish holotype proteins as defined by Crickmore et al. (1998). In this nomenclature, Cry proteins with similar amino acid sequences are grouped together. Cry proteins with the same Arabic numeral, e.g., Cry2, share at least 45% amino acid sequence identity. Those with the Arabic numeral and upper case letter, e.g., Cry2A, share at least 75% sequence identity. Finally, Cry proteins with the same Arabic numeral, upper case letter and lower case letter, e.g., Cry2Ab, share greater than 95% sequence identity. Bacillus thuringiensis (B.t.) is a gram-positive bacterium commonly present in soil and has been used commercially in the U.S. since 1958 in microbially derived products with insecticidal activity (U.S. EPA, 1988). Bacillus thuringiensis subsp. kurstaki, present in commercial microbial pest control products such as DiPel and Crymax, contain both the cry2aa and cry2ab2 genes. Although the Cry2Aa protein is produced in these commercial products, the cry2a2b gene is a pseudo gene, meaning that although the coding sequence is present, Cry2Ab protein is not produced due to an inefficient promoter (Dankocsik et al., 1990). Therefore, the Cry2Ab2 protein is not naturally produced in soil bacteria or sprayable microbial formulations (Widner and Whiteley, 1990; Crickmore et al., 1994). Both the cry2aa and cry2ab2 genes are located on the same 100 MDa plasmid (Donovan, et al., 1988; 1989) and the sequence of the cry2ab2 gene has been fully characterized (Widner and Whiteley, 1990). The Cry2Ab2 protein is derived from Bacillus thuringiensis, and is 88% amino acid sequence identical to the Cry2Aa protein produced by the B. thuringiensis kurstaki bacterium. This bacterial strain controls insect pests by the production of crystalline insecticidal proteins known as delta-toxins. Mode of Action and Specificity of the Cry2Ab2 Protein DiPel is a registered trademark of Abbott Laboratories. Crymax is a registered trademark of Ecogen, Inc. August

7 The Cry proteins exhibit a complex, multi-component mode of action. Insecticidal activity of the Cry proteins requires that the protein be ingested by the target insect pest. In the insect gut, the protein is solubilized due to the high ph of the insect gut and is proteolytically cleaved to the active core of the protein, which is resistant to further degradation by the insect gut proteases (Lilley et al., 1980; English and Slatin, 1992). The core protein binds to specific receptors on the mid-gut epithelium cells of susceptible insects, inserts into the membrane, and forms ion-specific pores (English and Slatin, 1992). The cells swell due to an influx of ions and water, leading to cell lysis and ultimately the death of the insect (Höfte and Whitely, 1989). The digestive tract tissues of non-target insects, mammals, birds, and fish do not contain receptors that bind the Cry proteins (Noteborn, 1994; Sacchi et al., 1986; Van Mellaert et al., 1988). Therefore, the Cry proteins cannot disrupt digestion in on-target species. Cry proteins are considered nontoxic to species other than lepidopteran and dipteran insects because there is a strong correlation between toxicity and specific binding of Cry proteins (Siegel et al., 2001; Betz et al., 2000; Hofmann et al., 1988). Characterization and History of Safe Consumption of Cry2Ab2 and GUS Proteins There is a history of safe use of Cry proteins in microbial B.t.-based products (U.S. EPA, 1998; IPCS, 1999). EPA and WHO have concluded that the potential dietary exposure to Cry proteins from use of microbial sprays on food crops does not raise any concerns: The use patterns for B. thuringiensis may result in dietary exposure with possible residues of the bacterial spores on raw agricultural commodities. However, in the absence of any toxicological concerns, risk from the consumption of treated commodities is not expected for both the general population and infants and children (U.S. EPA, 1998) and B.t. has not been reported to cause adverse effects on human health when present in drinking-water or food. (IPCS, 1999). The amino acid sequence of the Cry2Ab2 protein produced in Bollgard II cotton has been predicted based on nucleotide sequence of the coding sequence. The Cry2Aa protein exhibits a high degree of amino acid similarity (97%) with the 88% amino acid identical Cry2Ab2 protein produced in Bollgard II cotton. Thus, safety studies conducted with microbial B.t. products containing Cry2A proteins are relevant to the safety assessment of the Cry2Ab2 protein present in Bollgard II cotton. The Cry2A protein as a component of B.t. microbial products has been shown to have no deleterious effects on fish, avian species, mammals, and other non-target organisms (U.S. EPA, 1998; Betz et al., 2000). The lack of acute toxicity of the Cry proteins to non-target species is attributed to the highly specific mode of action and rapid digestibility. The GUS protein produced in Bollgard II cotton has an extensive history of safe use. Exposure of humans to the GUS protein is common, because GUS is present in intestinal epithelial cells, intestinal microflora bacteria, and numerous foods, and no harmful effects have been reported (Gilissen et al, 1998). GUS activity has been detected in over 50 plant species (Hu et al., 1990). These species include a number of human food sources, including potato, apple, almond, rye, rhubarb, and sugar beet (Schulz and Weissenbock, 1987; Hodal et al., 1992; Wozniak and Owens, 1994). GUS is also present in beef and in a August

8 number of invertebrate species, including nematodes, mollusks, snails, and insects (Gilissen et al., 1998). Even when ingested in raw foods such as shellfish or apples, GUS is not known to cause harmful effects (Gilissen et al., 1998). Likewise, the metabolites of E. coli-derived GUS are non-toxic (Gilissen et al., 1998). The E. coli-derived GUS enzyme produced by Bollgard II cotton is 99.8% homologous and functionally equivalent to the GUS enzyme from E. coli naturally present in the human gut. Digestion of Cry2Ab2 and GUS Proteins in Simulated Gastric and Intestinal Fluids In vitro, simulated mammalian gastric and intestinal digestive mixtures were used to assess the susceptibility of the Cry2Ab2 and GUS proteins to proteolytic digestion. Rapidly digested proteins represent a minimal risk of conferring novel toxicity or allergy, comparable to other safe dietary proteins (Astwood et al., 1996; Astwood and Fuchs, 2000). The rate of degradation of the Cry2Ab2 and GUS proteins was evaluated separately in simulated gastric (pepsin, ph 1.2) and intestinal (pancreatin, ph 7.5) fluids. The method of preparation of the simulated digestion solutions used is described in the United States Pharmacopoeia (1995). The degradation of the Cry2Ab2 protein was assessed by SDS-PAGE, western blot analysis and insect bioassay. SDS-PAGE analysis of simulated gastric fluid (SGF) demonstrated that greater than 98% of the Cry2Ab2 protein was digested within 15 seconds and that no fragments >2kDa of the parent protein were resolved. The acid conditions of the stomach denature the native conformation of the Cry2Ab2 protein, facilitating its rapid digestion. Western blot analysis of simulated intestinal fluid (SIF), showed that within one minute the Cry2Ab2 protein was degraded to a relatively stable protein fragment ( 50kDa) that was bioactive for at least 24 hours. This result was expected because protease-resistant core proteins of B.t. insecticidal proteins are known to be resistant to further trypsin digestion (Lilley et al., 1980). In vivo, the Cry2Ab2 protein would be exposed to gastric conditions prior to entering the intestinal lumen. The low ph and pepsin in the stomach would be expected to either fully digest the protein or cause it to become susceptible to intestinal digestion. The degradation of the GUS protein was assessed by western blot analysis and enzymatic activity assays. Within 15 seconds of exposure to SGF, there was no detectable GUS protein in either assay. After two hours in SIF, 91% of the original GUS activity was lost in the enzyme assay, with only a faint band detected in the western blot analysis. Based on these results, it was concluded that any GUS protein ingested by humans would be readily degraded in the digestive tract (Fuchs and Astwood, 1996). Human exposure to either the Cry2Ab2 or the GUS protein from cotton-derived products would not be expected because cotton processing removes or denatures both the Cry2Ab2 and GUS proteins (refer to section below on Assessment of Human Exposure to Cry2Ab2 and GUS Proteins from Bollgard II Cotton). Assessment of Acute Oral Toxicity of Cry2Ab2 and GUS Proteins in Mice August

9 Few proteins are toxic when ingested. Those that are toxic, typically act acutely (Sjoblad et al., 1992). Results of mammalian acute oral toxicity studies of Cry2Ab2 and GUS proteins support their specificity and lack of acute toxicity. There was no evidence of toxicity even at extremely high dose levels. There were no treatment-related adverse effects in mice administered Cry2Ab2 protein by oral gavage at doses up to 1450 mg/kg of body weight. Similarly, the GUS protein caused no deleterious effects when administered by oral gavage at doses up to 69 mg/kg. Results from these studies demonstrated that the Cry2Ab2 and GUS proteins are not acutely toxic to mammals. This result was expected because both the Cry2Ab2 and GUS proteins are readily digested in gastric and intestinal fluids in vitro and both proteins are from families of proteins with a history of safe consumption. Assessment of Sequence Similarity of Cry2Ab2 and GUS Proteins to Known Protein Toxins Another aspect used for the assessment of potential toxic effects of proteins introduced into plants is to compare the amino acid sequence of the protein to sequences of known toxic proteins. Homologous proteins derived from a common ancestor have similar amino acid sequences, are structurally similar and share common function. Therefore, it is undesirable to introduce DNA that encodes a protein that is homologous to any toxin. Homology is determined by comparing the degree of amino acid similarity between proteins using published criteria (Doolittle et al., 1990). The Cry2Ab2 protein does not show meaningful amino acid sequence similarity when compared to known protein toxins present in the PIR, EMBL, SwissProt, and GenBank protein databases, with the exception of other Cry proteins. The GUS protein does not show any meaningful amino acid sequence similarity when compared to known protein toxins present in these protein databases. Assessment of Potential Allergenicity of Cry2Ab2 and GUS Proteins Although there are no single predictive bioassays available to assess the allergenic potential of proteins in humans (U.S. FDA, 1992), the physicochemical and human exposure profile of the protein provides a basis for assessing potential allergenicity by comparing it to known protein allergens. Thus, important considerations contributing to the allergenicity of proteins ingested orally include exposure and an assessment of the factors that contribute to exposure, such as stability to digestion, prevalence in the food, and consumption patterns (amount) of the specific food (Metcalfe et al., 1996; Kimber et al., 1999). A key parameter contributing to the systemic allergenicity of certain food proteins appears to be stability to gastrointestinal digestion, especially stability to acid proteases like pepsin found in the stomach (Astwood et al., 1996; Astwood and Fuchs, 1996; Fuchs and Astwood, 1996; FAO, 1995; Kimber et al., 1999). Important protein allergens tend to be stable to peptic digestion and the acidic conditions of the stomach if they are to reach the intestinal mucosa where an immune response can be initiated. As noted above, the in vitro assessment of the Cry2Ab2 and GUS proteins digestibility indicates that these proteins are readily digested. August

10 Another significant factor contributing to the allergenicity of certain food proteins is their high concentration in foods (Taylor et al., 1987; Taylor, 1992; Fuchs and Astwood, 1996). Most allergens are present as major protein components in the specific food, representing from 2-3% to as high as 80% of total protein (Fuchs and Astwood, 1996). In contrast, the Cry2Ab2 and GUS proteins are present at relatively low levels in Bollgard II cotton plants. The Cry2Ab2 and GUS proteins represent approximately <0.004% and 0.007%, respectively, of the total protein in seed. It is also important to establish that the protein does not represent a previously described allergen and does not share potentially cross-reactive amino acid sequence segments or structure with a known allergen. An efficient way to assess whether the added protein is an allergen or is likely to contain cross-reactive structures is to compare the amino acid sequence with that of all known allergens. A database of protein sequences associated with allergy and coeliac disease has been assembled from publicly available genetic databases (GenBank, EMBL, PIR, and SwissProt). The amino acid sequences of the Cry2Ab2 and GUS proteins were compared to these sequences. Neither the Cry2Ab2 protein nor the GUS protein shares any meaningful amino acid sequence similarity with the known allergens (Astwood et al., 1996). In summary, these data and analyses support the conclusion that Cry2Ab2 and GUS proteins do not pose a significant allergenic risk, are not derived from allergenic sources, do not possess immunologically-relevant sequence similarity with known allergens, and do not possess the characteristics of known protein allergens. In addition, as discussed in the next section, these proteins are not detectable in cotton products used for human food. Characteristics of known allergenic proteins Characteristic Allergens 1 Cry2Ab2 GUS Stable to digestion yes no no Stable to processing yes no no Similarity to known allergens yes no no Prevalent protein in food yes no no 1 As described in Taylor (1992) and Taylor et al. (1987) Assessment of Human Exposure to Cry2Ab2 and GUS Proteins in Bollgard II Cotton Cottonseed oil and processed cotton linters are the major cotton products used for human food (National Cottonseed Products Association, 1989). Analysis of refined cottonseed oil and processed cotton linters derived from both the parental Coker control line and Bollgard cotton event 531 confirmed that there is no detectable protein in cottonseed oil (detection August

11 limit = 1.3 ppm total protein). This is consistent with other reports that conclude that there is an absence of protein in cottonseed oil (Cottonseed Oil, 1993). Analysis of processed linters also confirmed there is no detectable protein (Sims et al., 1996). Likewise, there is no reason to expect that the Cry2Ab2 or the GUS proteins would be present in cottonseed oil or linters of Bollgard II cotton. Therefore, significant human consumption of the Cry2Ab2 and GUS proteins present in Bollgard II cotton varieties is extremely unlikely. Furthermore, direct food challenge of individuals allergic to proteins contained in the meal derived from oilseed crops (e.g., soybean, peanut, and sunflower) with the oil from these respective crops has established that refined oil does not elicit an allergenic response (Bush et al., 1985; Halsey et al., 1986; Taylor et al., 1981). This lack of response is consistent with the lack of detectable protein in the oil (Tattrie and Yaguchi, 1973). This information supports the conclusion that there is insignificant human exposure to the Cry2Ab2 and GUS proteins in Bollgard II cotton, and that Bollgard II cottonseed oil poses no significant allergenic concerns. Compositional Analysis and Nutritional Assessment of Bollgard II Cotton The design of a food and feed safety assessment program for a genetically engineered crop requires detailed understanding of the uses of the crop and crop products in animal and human nutrition. Cotton is the leading plant fiber crop produced in the world and is grown primarily for its fiber. Cottonseed is processed to produce animal feed ingredients. Cottonseed meal is primarily used as cattle feed, with smaller proportions of meal fractions used in feed for poultry, sheep, catfish, and swine. Cottonseed serves as an excellent source of fiber and protein in animal feed, particularly due to its high lysine content. Oil is the main food ingredient derived from cottonseed and is used for frying oil and in salad dressings. Compositional Analysis To assess whether the composition of Bollgard II cotton is comparable to conventional cotton present in the marketplace, with the exception of the introduced trait, compositional analyses were performed on the cottonseed from Bollgard II cotton event 15985, the DP50B parental variety, the DP50 non-transgenic control variety, and commercial cotton varieties produced in 1998 from eight locations within six states in the U.S. (Texas, Arizona, Mississippi, South Carolina, Louisiana, and Alabama). Forty-eight different compositional components were evaluated. These analyses included: Proximate analysis: protein, fat, ash, water, carbohydrate, calories (Table 2); Amino acid composition: levels of individual amino acids (Table 3); Fatty acid profile: total lipid content and percentage of individual fatty acids in raw seed (Table 4) and refined cottonseed oil (Table 5); Minerals: calcium, copper, iron, magnesium, manganese, phosphorus, potassium, sodium, and zinc (Table 6); Anti-nutrients: levels of gossypol, cyclopropenoid fatty acids, and aflatoxins in seed (Table 7); levels of gossypol in refined oil and cottonseed meal (Table 8); and cyclopropenoid fatty acids in refined oil (Table 9). August

12 Statistical analyses were conducted on the data using a mixed model analysis of variance for a combination of all sites for The results of these analyses, summarized in Tables 3 to 9, demonstrate that seed from Bollgard II cotton is compositionally equivalent to seed from the DP50B parent variety and other commercial cotton varieties. In the 50 comparisons, there were six instances where the mean values of Bollgard II cotton were statistically significantly different from the parental line (Table 10). In all of these instances, the means were within the range of levels found for commercial cotton. Furthermore, the statistical differences were not observed consistently at all locations and as a result were not considered to be biologically relevant. Fatty acid profiles were evaluated in cottonseed for Bollgard II cotton and there were no statistically significant differences in palmitic, palmitoleic, oleic, linolenic, and gamma linoleic, arachidic, behenic, or lignoceric acids compared to DP50B (Table 4). Small but statistically significant differences were observed for myristic, stearic, and linoleic acids between Bollgard II cotton and the DP50B control. All significantly different mean values for Bollgard II cotton were within the non-transgenic and commercial cotton reference ranges (Table 10), as well as within the ranges published in the literature (Berberich et al., 1996). Therefore, these differences were not considered biologically relevant. Mineral levels were measured in cottonseed (Table 6). There were no statistically significant differences in mineral levels obtained for Bollgard II cotton compared to the DP50B control and the means were within the non-transgenic and commercial reference ranges. Levels of anti-nutrients contained in cottonseed, such as gossypol, were comparable for Bollgard II cotton and the parental cotton line. The primary aflatoxins (B1, B2, G1, G2) were undetected in the Bollgard II, DP50B control, and the reference cotton lines at an LOD of 0.1 ppb (Table 7). Statistically significant differences were observed for the mean values of the cyclopropenoid fatty acids, malvalic, diydrosterculic, and sterculic acids, in comparisons of values for Bollgard II cotton to the parental control (Table 10). All the significantly different comparisons of mean values were within the ranges for the parental and commercial reference ranges, as well as literature ranges. Additionally, only one of the four replicated field locations showed the statistically significant differences in the mean comparisons of Bollgard II cotton to the control. Therefore, these differences were not considered biologically relevant. The major cottonseed processed products, refined oil and meal, were also shown to be equivalent to those products produced from the control cotton line. The refined oil was evaluated for fatty acid profile, free gossypol content, and cyclopropenoid fatty acid levels. The fatty acid profile of the refined oil was typical of commercial cottonseed oil (Table 5). Free gossypol was not detectable and cyclopropenoid fatty acid levels were similar to levels previously reported in the literature for both cottonseed meal and oil (Table 8 and 9). The full fat flour and toasted meal were analyzed for total gossypol levels. When cottonseed is flaked and heated during processing to oil and meal, the cotton lysigenous glands are ruptured and gossypol is released. Some of the gossypol binds to seed August

13 components, primarily to proteins through the free amino groups of lysine. The binding of gossypol during processing is important because the free form of gossypol is considered toxic, whereas the bound form is unavailable and essentially inactive (Martin, 1990; Berardi and Goldblatt, 1980). As expected, there was no detectable free gossypol in toasted meal (Table 8). Therefore, insertion of the cry2ab2 and uida genes in the cotton genome did not alter the processing characteristics of the cottonseed. In summary, the results of numerous analytical measurements of composition demonstrate that Bollgard II cottonseed is compositionally equivalent to the parental variety and conventional cotton varieties. Processing is unlikely to alter the compositional components of cotton and, therefore, products derived from cottonseed will also be compositionally equivalent to and as safe as current cotton-derived products. Nutritional Assessment and Toxicological Assessment of Cottonseed In addition to the compositional studies, the nutritional wholesomeness of seed from Bollgard II cotton was demonstrated by feeding rats, channel catfish, and dairy cows diets which contained cottonseed from both the Bollgard cotton and control cotton cultivars. At completion of the rat feeding study, there were no significant differences in weight gain or feed intake between rats consuming Bollgard II cotton and the control cotton diet. Similarly, there were no significant differences in survival, weight gain, feed conversion ratio, or fillet composition between channel catfish fed a diet containing Bollgard cotton compared to catfish fed the control cottonseed diet. Results of a cow study also showed that cottonseed of Bollgard cotton is as wholesome and nutritious as control cottonseed for cows based on similar feed intakes, general health and milk production and composition (Castillo et al., 2001). Results of these studies confirm the food and feed safety and nutritional equivalence of diets from Bollgard II cotton event to diets from conventional cotton varieties. Horizontal Gene Transfer and the Assessment of Marker Genes Horizontal gene transfer is defined as the transfer of DNA from one species to another. With respect to crop plants that are developed through biotechnology, a number of assessments have been performed to evaluate the possibility that antibiotic resistance marker genes used to facilitate the selection of the transformed plants might be transferred to bacteria either in the field or in animals that have consumed the crop. The reason for the assessment is that some species of bacteria found in soil, in the rumen or in the intestine can receive DNA from other organisms through three mechanisms of transfer (Morrison, 1996; Davison, 1999). However, transformation is the only relevant mechanism to the possible transfer of DNA from plants to bacteria and subsequent expression of the encoded protein product. The other two mechanisms, conjugation (exchange of plasmid DNA between compatible bacteria) and transduction (viral transfer of DNA into bacteria) are specific to restricted forms of transfer and are not relevant to the potential transfer of DNA from plants (Thomson, 2000). August

14 Because Bollgard II cotton is a re-transformation of Bollgard cotton, the assessment of possible horizontal gene transfer of the antibiotic resistance genes between Bollgard II cotton and other organisms producing antibiotic resistance marker genes and microorganisms has previously been discussed (Monsanto, 2002). In general, bacterial species differ markedly in their ability to accept DNA from the environment, and the frequency of transformation, even under ideal circumstances, is very low. The DNA that was transferred into cotton to produce Bollgard II cotton was incorporated into the genomic DNA of the plant and represents a small fraction of the cotton genome. For Bollgard II cotton, the origin of replication for plasmid maintenance at high copy number in E. coli, ori322, was contained on the plasmid PV-GHBK11 used for transformation, but was not transferred into the cotton plant genome. Therefore, the antibiotic resistance genes in Bollgard II cotton cannot be mobilized by excision of the marker gene to create a functional plasmid. The DNA would have to be integrated into the recipient s genome or plasmid in order to replicate and be passed on through reproduction. Studies have addressed this potential for the horizontal transfer of antibiotic selectable marker genes and concluded the probability of this event occurring is virtually zero (Prins and Zadoks, 1994; Schlüter et al., 1995; Nielsen et al., 1998; Beever and Kempe, 2000; Jelenić, 2003). Environmental Assessment Cotton Cotton is of the genus Gossypium, of the tribe Gossypieae, and of the family Malvaceae. Worldwide, four species of cotton are of agronomic importance: the two diploid Asiatic species, G. arboreum and G. herbaceum, and the two-allotetraploid New World species, G. barbadense and G. hirsutum. Although the diploid species remain important in restricted areas of India, Asia, and Africa, the two New World species account for approximately 98% of world cotton fiber production. Wild species of Gossypium typically occur in arid parts of the tropics and sub-tropics. Wild populations of G. hirsutum are relatively rare and tend to be widely dispersed. All grow on beach strands or on small islands. Cotton is normally considered a self-pollinating crop but can be cross-pollinated by certain insects. However, outcrossing of the cry2ab2 gene from Bollgard II cotton to other Gossypium species or to other Malvaceous genera is extremely unlikely for the following reasons (Percival et al., 1999): cultivated cotton is an allotetraploid and is incompatible with cultivated or wild diploid cotton species; therefore, it cannot cross and produce fertile offspring. although outcrossing to wild or feral allotetraploid Gossypium species can occur, cotton production generally does not occur in the same geographical locations as the wild relatives. For example, outcrossing to G. tomentosum in Hawaii is possible, but cotton is not grown commercially in Hawaii. there are no known plant species other than those of the genus Gossypium that are sexually compatible with cultivated cotton. August

15 If the cry2ab2 gene were transferred to a wild population of a tetraploid cotton species, and if this was considered undesirable, the size of the plants, their perennial growth habit, their restricted habitat, and their low natural fecundity would make them easy to control. Crossing of insect protection genes to other cultivated cotton genotypes is possible should the plants be in close proximity; however, studies have shown that this occurs at a very low frequency and is not considered to be a concern as it is unlikely to cause any adverse impact to the environment (Green and Jones, 1953; Mehetre, 1992). Assessment of Agronomic Performance Bollgard II cotton has been grown and observed at multiple locations for weediness, plant growth characteristics, susceptibility to insects, and disease infection. Based on results of the field monitoring programs, there were no significant differences in agronomic characteristics between Bollgard II cotton and the parental DB50 variety. Bollgard II cotton does not pose any different plant pest risk to other plants and the environment than non-transformed cotton varieties. Bollgard II cotton meets all morphological, yield, and quality characteristics of cotton varieties produced in the United States. Cotton is not considered to have weedy characteristics as an annual plant grown in the United States. It does not possess any of the attributes commonly associated with weeds such as seed dormancy, long soil persistence, germination under diverse environmental conditions, rapid vegetative growth, a short life cycle, high seed output, high seed dispersal, or long distance dispersal of seeds. Multiple genes typically control these characteristics of weeds. Wild populations of cotton are rare, widely dispersed and confined to beach strands or to small islands (Lee, 1984). Cotton appears to be somewhat opportunistic towards disturbed land and is not especially effective in invading established ecosystems. There is little probability that Bollgard II cotton or any Gossypium species crossing with Bollgard II cotton could become a weed. All wild and feral relatives of cotton are tropical, woody, perennial shrubs (Percival et al., 1999), other than a few herbaceous perennials in northwest Australia. With the exception of G. thurberi and G. sturtianum in Australia, these cannot naturally exist even in the milder temperate regions. In most instances the distribution of these species is determined by soil and climatic conditions. As perennials, the plants do not tend to produce seed each year. In fact, they tend to drop fruit in response to stress. It is unlikely that production of the Cry2Ab2 protein would impact survival either way. Bollgard II cotton does not have any different weediness characteristics than other conventional cotton varieties. Bollgard II cotton does not exhibit different agronomic or morphological traits compared to controls, which would confer a competitive advantage over other species in the ecosystem in which it is grown. Based on these mechanistic arguments and field experience, there is no indication that insertion of the cry2ab2 gene into the cotton genome would have any effect on the weediness traits of the cotton plant. August

16 Assessment of Effect to Non-Target Organisms There is extensive information about microbial preparations of Bacillus thuringiensis subsp. kurstaki (B.t.k) containing Cry proteins that demonstrate that these proteins are nontoxic to non-target organisms (U.S. EPA, 1988; Betz et al., 2000). The literature has established that the Cry proteins are extremely selective for lepidopteran insects, bind specifically to receptors on the mid-gut of lepidopteran insects, and have no deleterious effect on beneficial/non-target insects (Hofmann et al., 1988; English and Slatin, 1992; Betz et al., 2000; Siegel et al., 2001). To confirm and expand on results obtained for the microbial products, the potential impact of the Cry2Ab protein on non-target organisms was assessed on several representative organisms. The non-target organism species included larval and adult honey bee (Apis mellifera L.), a beneficial insect pollinator; green lacewing larvae (Chrysopa carnea), a beneficial predaceous insect commonly found on cotton and other cultivated crops; parasitic Hymenoptera (Nasonia vitripennis), a beneficial parasitic wasp of the housefly; the ladybird beetle (Hippodamia convergens), a beneficial predacious insect which feeds on aphids and other plant bugs commonly found on stems and foliage of weeds and cultivated plants; Collembola (Folsomia candida) and earthworm (Eisemia fetida) nontarget soil organisms; and northern bobwhite quail (Table 11). No adverse effects were observed at the maximum expected environmental concentrations to which these non-target organisms would be exposed. In all studies conducted, a NOEC (no observed effect concentration) was established and found to exceed predicted maximum environmental concentrations. In most studies, the NOEC exceeded the maximum predicted environmental concentration by 10- to over 100-fold, demonstrating a wide margin of safety for these organisms. In summary, Cry proteins exhibit a high degree of specificity and therefore do not pose a significant hazard to non-target animals such as mammals, birds, fish, water fleas, earthworms, and beneficial insects. Although several endangered lepidopteran and dipteran species may potentially be susceptible to Cry proteins, no exposure is predicted because of their feeding habit or because the habitats of these endangered species in cottongrowing areas do not overlap with cotton fields. Fate of Cry2Ab2 Protein in Soil The results of a soil degradation study demonstrate that the Cry2Ab2 protein dissipates rapidly in the soil environment. Analysis of soils from Mississippi, Arizona and Alabama treated with purified Cry2Ab2 protein by insect bioassay established a DT 50 range of days; the DT 90 range was days. These results support the conclusion that the Cry2Ab2 protein derived from Bollgard II cotton degrades rapidly in soil. In addition, the short DT 50 and DT 90 values obtained in soils dosed with a solution of pure Cry2Ab2 protein suggest that any Cry2Ab2 protein that would reach soil as the pure protein (e.g., by root exudation, or otherwise not combined with tissue) would be degraded in less than 6 days. August

17 The rapid degradation of the Cry2Ab2 protein in soil ensures exposure risk for soil dwelling organisms will be minimal. Assessment of Genetic Stability The cry2ab2 gene conferring insect protection in Bollgard II cotton event was demonstrated as stably integrated into the chromosome. This conclusion is based on molecular analyses, data on phenotypic expression, and inheritance patterns. The results of these studies are summarized as follows: molecular analyses of plants from the R3 to R5 generations establish that the introduced genes are maintained in the same chromosomal location; analyses of seed obtained from multi-site trials using R4 and R5 generations showed no marked change in production of Cry2Ab2 protein; the level of insect protection has been maintained for at least six generations and during testing in the US over the last six years under diverse environmental conditions and in many cotton lines with different genetic backgrounds; production of Cry2Ab2 protein has been confirmed under different environmental conditions and in many cotton lines with different genetic backgrounds; Mendelian inheritance of the Cry2Ab2 protein production is observed after selfpollination or backcrossing with other cotton varieties; and seed quality (germination, vigor) of Bollgard II cotton is maintained after transfer of the cry2ab2 gene into cotton from different genetic backgrounds. In summary, it is concluded that the inserted genes in Bollgard II cotton event are stably integrated and the line is phenotypically and genetically stable over several generations, and in various environments. Insect Resistance Management Effective insect resistance management (IRM) programs for B.t. crops are a vital part of responsible product stewardship and should be instituted based on the best available knowledge, employing what is known about the trait, the mode of action, the targeted insects and the environment in which the product is introduced, while being properly respectful of uncertainties so as to make B.t. technologies available to growers as an additional pest management tool. Such programs must strike a balance between available knowledge and practicality, with grower acceptance and implementation of the plan as critical components. Monsanto supports the development and implementation of an effective and practical IRM plan for all B.t. crops in all markets where these products are introduced. Each plan includes the following elements: Baseline susceptibility determination for the target pests and surveillance for changes in susceptibility; An adequate supply of susceptible insects to mate with any resistant insects (achieved through appropriate practical programs such as structured refuge, natural or cultivated alternate hosts, grower practices, etc.); August

18 Mitigation plans; and Grower awareness, education of IRM concepts, and some means of assessing grower behavior when particular IRM practices are required of them. These plans vary according to geography, pest and overlapping crops, and are reviewed on a regular basis with updated information available from interested stakeholders. Combining the Cry2Ab2 protein with the Cry1Ac protein already in the marketplace further strengthens the IRM strategy for Bollgard II cotton. The second insecticidal protein in Bollgard II cotton is sufficiently different in its mechanism of action from the Cry1Ac and is highly efficacious against the target pest species. Insects would need to develop two distinct modes of resistance to survive both proteins, which is highly unlikely. Therefore, if a target insect should develop resistance to one of the proteins, the other protein will provide control of that resistant insect. Environmental Assessment Conclusions In summary, comparisons of Bollgard II cotton event plants were made to conventional cotton plants with regard to disease and pest characteristics, yield, morphology, weediness, impact on non-target organisms, and other characteristics. Based on results of these extensive studies, it was concluded that the trait for protection from lepidopteran insect pests is stably inherited and that Bollgard II cotton event does not pose any increased plant pest risks or environmental risks compared to conventional cotton varieties. Furthermore, the combination of the Cry2Ab2 and the Cry1Ac proteins provide an enhanced IRM strategy to delay the development of resistance in lepidopteran insects. Summary Bollgard II cotton, which has two modes of action for improved lepidopteran control and increased spectrum of activity over Bollgard cotton, is expected to provide significant benefits to cotton production including the reduction in pesticide use, improved control of target insect pests, improved yield, reduced production costs, and improved profitability for cotton growers. Detailed food, feed, and environmental safety assessments confirm the safety of this product. The analyses included: 1) detailed molecular characterization of the introduced DNA; 2) safety assessments of the produced Cry2Ab2 and GUS proteins; 3) compositional analysis of cottonseed, oil, and meal; and 4) environmental impact assessment of the cotton plants. These studies demonstrate that the Cry2Ab2 protein is safe to non-target organisms, including humans, animals, and beneficial insects. Additionally, Bollgard II cotton plants and cottonseed were shown to be as safe and nutritious as conventional cotton varieties. Information and data contained within this document have been provided to regulatory authorities for review. Regulatory review continues as we update regulatory files and make submissions to additional countries globally. August

19 References Astwood, J.D. and R.L. Fuchs Food allergens are stable to digestion in a simple model of the gastrointestinal tract. Journal of Allergy and Clinical Immunology. 97:241. Astwood, J.D. and R.L. Fuchs Status and safety of biotech crops. Pp In Agrochemical discovery insect, weed and fungal control. Baker D.R. and N.K. Umetsu (eds.). ACS Symposium Series 774. Astwood, J.D., J.N. Leach, and R.L. Fuchs Stability of food allergens to digestion in vitro. Nature Biotechnology. 14: Baum, J.A. T.B. Johnson, and B.C. Carlton Bacillus thuringiensis natural and recombinant bioinsecticide products. In Methods in Biotechnology. Pp Vol 5. Biopesticides: Use and Delivery (F.R. Hall and J.J. Mean (eds). Humana Press, Inc., Totowa, N.J. Beever, D.E., and C.F. Kemp Safety issues associated with the DNA in animal feed derived from genetically modified crops. A review of scientific and regulatory procedures. Nutrition Abstracts and Reviews Series B: Livestock feeds and feeding. 70: Berardi, L.C. and L.A. Goldblatt In Toxic Constituents of Plant Foodstuffs. 2nd Ed., I.F. Liener, ed., Academic Press, New York, pp Berberich, S.A., J.E. Ream, T.L. Jackson, R. Wood, R. Stipanovic, P. Harvey, S. Patzer, and R.L. Fuchs The composition of insect-protected cottonseed is equivalent to that of conventional cottonseed. J Ag Food Chem. 44: Betz, F.S., B.G. Hammond and R.L. Fuchs Safety and advantages of Bacillus thruringiensis-protected plants to control insect pests. Regulatory Toxicology and Pharmacology. 32: Bush, R.K., S.L. Taylor, J.A. Nordlee, and W.W. Busse Soybean oil is not allergenic to soybean-sensitive individuals. J. Allergy Clin. Immunololgy. 76: Carpenter, J.E. and L.P. Gianessi Agricultural biotechnology: updated benefit estimates. Pp National Center for Food and Agricultural Policy, Washington, D.C. Castillo, A.R., M.R. Gallardo, M. Maciel, J.M. Giordano, G.A. Conti, M.C. Gaggiotti, O. Quaino, C. Gianni, and G.F. Hartnell Effect of feeding dairy cows with either Bollgard Bollgard II, Roundup Ready or control cottonseeds on feed intake, milk yield and milk composition. J. Dairy Sci., Vol 48, suppl. 1, Abstract August

20 Cottonseed Oil L.A. Jones and C.C. King (ed.), National Cottonseed Products Association, Inc. and the Cotton Foundation, Memphis, TN. 60pp. Crickmore, N., V.C. Wheeler, and D.J. Ellar Use of an operon fusion to induce expression and crystallization of a Bacillus thuringiensis delta-endotoxin encoded by a cryptic gene. Mol. Gen. Genet. 242: Crickmore, N., D.R. Zeigler, J. Feitelson, E. Schnepf, J. Van Rie, D. Lereclus, J. Baum, and D.H. Dean Revision of the nomenclature for the Bacillus thuringiensis pesticidal crystal proteins. Microbiol. Mol. Biol. Rev. 62: Dankocsik, C., W.P. Donovan, and C.S. Jany Activation of a cryptic crystal protein gene of Bacillus thuringiensis subspecies kurstaki by gene fusion and determination of the crystal protein insecticidal specificity. Molecular Microbiology 4: Davison, J Genetic exchange between bacteria in the environment. Plasmid 42: Donovan. W.P., C. Dankocsik, M.P. Gilbert, M.C. Gawron-Burke, R.G. Groat, and B.C. Carlton Amino acid sequence and entomocidal activity of the P2 crystal protein, an insect toxin from Bacillus thuringiensis var. kurstaki. J. Biol. Chem. 263: Donovan. W.P., C. Dankocsik, M.P. Gilbert, M.C. Gawron-Burke, R.G. Groat, and B.C. Carlton Additions and Corrections. J. Biol. Chem. 264:4740. Doolittle, R.F., D.F. Feng, K.L Anderson, and M.R. Alberro A naturally occurring horizontal gene transfer from a eukaryote to a prokaryote. J. of Molecular Evolution. 31: Economic Research Service/USDA Genetically engineered crops: has adoption reduced pesticide use? Agricultural Outlook. August, Pp Edge, J. M., J.H. Benedict, J.P. Carroll, and H.K. Reding Bollgard Cotton: An assessment of global economic, environmental, and social benefits. J. Cotton Science. 5:1-8. English, L. and S.L. Slatin Mode of action of delta-endotoxins from Bacillus thuringiensis: A comparison with other bacterial toxins. J. Biochem. Molec. Biol. 22(1):1-7. Falck-Zepeda, J.B., G.Traxler, and R.G. Nelson Rent creation and distribution from the first three years of planting Bt cotton. Agribusiness. 16:1-25. ISAAA Briefs No. 14. ISSA, Ithaca, N.Y. August