Journal of INVERTEBRATE PATHOLOGY Journal of Invertebrate Pathology 95 (2007) 227–230 www.elsevier.com/locate/yjipa

Insect resistance management for Syngenta’s VipCotä transgenic

Ryan W. Kurtz a,*, Alan McCaffery b, David O’Reilly a

a Syngenta , Inc., Research Triangle Park, NC 27709, USA b Syngenta, Jealotts Hill Research Centre, Bracknell, Berks., RG42 6EY, UK

Received 28 January 2007; accepted 17 March 2007 Available online 25 March 2007

Abstract

Syngenta is seeking commercial registration for VipCotä cotton, a pyramided transgenic cotton trait that expresses two insecticidal proteins derived from Bacillus thuringiensis Vip3A and Cry1Ab. Both proteins are highly effective against two key cotton pests, Heli- coverpa zea cotton bollworm; and Heliothis virescens, budworm. To investigate the role of VipCotä cotton in delaying the devel- opment of resistance in these pests to transgenic Bt traits, Syngenta has performed studies to determine the dose of proteins expressed in VipCotä and evaluate the potential for cross-resistance between the component proteins. Following Environmental Pro- tection Agency (US EPA) high dose methods 1 and 4, VipCotä was shown to express a high dose of proteins for H. zea and H. virescens. VipCotä was also confirmed to express a high dose of proteins for H. zea through US EPA Method 5. Additionally, all the data collected to date verify a lack of cross-resistance between Vip3A and Cry proteins. These two key pieces of information indicate that VipCotä cotton should be very durable under the currently mandated high dose plus refuge insect resistance management strategy. Ó 2007 Elsevier Inc. All rights reserved.

Keywords: Bacillus thuringiensis; Insect resistance management; Vip3A; Helicoverpa zea; Heliothis virescens

1. Introduction protectants (PIP). The high dose refuge strategy involves presenting an insect with a toxin dose high enough to kill Syngenta’s VipCotä cotton was developed through con- all or nearly all insects heterozygous for a resistance allele, ventional breeding of two independent transgenic events, in combination with ensuring an appropriate source of COT102 cotton, which produces the Vip3A vegetative insects that have not been selected for resistance. The unse- insecticidal protein, and COT67B cotton, which produces lected insects originate from plants that do not express the a full-length version of the CrylAb endotoxin protein. Both toxin (i.e. a refuge). The numerous homozygous susceptible proteins expressed by VipCotä cotton are derived from insects produced from the refuge will mate with any rare Bacillus thuringiensis (Bt) and are active against a wide resistant insects surviving on the PIP giving rise to range of lepidopteran pest insects including the primary heterozygous offspring which are then killed by the high targets Helicoverpa zea (Boddie) and Heliothis virescens dose of toxin expressed by PIP (Roush, 1997, 1998). (F.). The first commercially available transgenic insect- Despite the effectiveness of expressing single insec- resistant crops express a single Cry1A insecticidal protein. ticidal proteins in delaying resistance development under Through Scientific Advisory Panels held in 1998 and 2001, the high dose refuge strategy, it is possible to develop a (US EPA, 1998; US EPA, 2001) the US EPA determined product with greater insect control characteristics and that a high dose plus refuge strategy should be the basis exceptional benefits from an IRM perspective by combin- for the insect resistance management (IRM) strategy to ing two insecticidal proteins that do not share cross-resis- delay resistance development to these plant incorporated tance in the same plant (Roush, 1998; Caprio, 1998). Here we provide an overview of the IRM data collected * Corresponding author. Fax: +1 919 541 8585. in support of commercial registration in the United States E-mail address: [email protected] (R.W. Kurtz). for Syngenta’s VipCotä cotton trait. These data include

0022-2011/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jip.2007.03.014 228 R.W. Kurtz et al. / Journal of Invertebrate Pathology 95 (2007) 227–230 the dose of the combined proteins expressed in VipCotä 100 Coker 312 4.00% 90 and evidence supporting a lack of cross-resistance between Coker 312 0.80% Vip3A and Cry toxins. 80 VipCot™ 4.00% 70 VipCot™ 0.80% 60 2. Studies to determine the high dose status of VipCotä 50

Mortality 40 cotton

% 30 20 When evaluating the potential effectiveness of an IRM 10 plan for a PIP derived from Bt, the US EPA requires data 0 on the relative insecticidal activity, or dose, the plant pro- Test 1 Test 2 Test 3 Test 4 Mean vides of the incorporated proteins. The US EPA defined a Fig. 2. Bioassay of H. zea on lyophilized Coker312 and VipCotä leaf high dose as 25 times the protein concentration needed to tissue. Neonate H. zea larvae were exposed to 4 and 0.8% preparations of kill susceptible larvae and recognizes five equal yet imper- lyophilized Coker 312 and VipCotä leaf tissue. Other details are as fect methods to demonstrate that a transgenic crop described in the legend to Fig. 1. expresses a high protein dose (US EPA, 1998, 2001). Of the five available methods, the criterion of two methods Exposure of the insects to 0.8% VipCotä leaf tissue must be met to demonstrate a high dose. Syngenta chose resulted in 97.3% mean mortality. Only low levels of mor- three of these methods with which to demonstrate the dose tality were observed among insects exposed to the control status of VipCotä. The methodology for Methods 1 and 5 Coker312 tissue. This meets the criterion for demonstrating is completely described in O’Reilly (2007) and Method 4 is a high dose for VipCotä against H. virescens following US briefly described below. EPA Method 1. Fig. 2 presents similar data from the bioassay of lyoph- 2.1. Method 1: Lyophilized tissue bioassay ilized Coker312 and VipCotä leaf tissue against H. zea. Exposure of the insects to either a 4 or 0.8% suspension US EPA Method 1 involves the bioassay of lyophilized of VipCotä leaf tissue resulted in 100% mortality in every tissues from transgenic plants using tissues from non-trans- case. Again, only low mortality was observed on control genic plants as controls. By diluting the lyophilized tissue samples. This meets the criterion for demonstrating a high in an agar suspension and overlaying it on artificial insect dose for VipCotä against H. zea following US EPA diet, it is possible to reduce the exposure of a target insect Method 1. species by at least 25-fold (4% dilution of leaf tissue in diet) relative to the living plant. A result of 99.9% mortality of 2.2. Method 5: Bioassay of 4th instar H. zea neonate insects using this method demonstrates high dose. The results of the bioassays of lyophilized Coker312 and US EPA Method 5 involves the identification of an VipCotä leaf tissue against H. virescens are shown in instar that is at least twenty five-fold more tolerant to the Fig. 1. Data from four independent repeats of the test, toxin than neonates and determining whether the PIP and the mean data across all four tests are presented. 4% shows at least 95% mortality of insects from that older VipCotä leaf tissue caused 100% mortality in all tests. instar compared to neonates. The older instar insects serve as surrogate heterozygous neonates carrying a resistance 100 Coker 312 4% 90 Coker 312 0.80% 100 80 Coker 312 L1 VipCot™ 4% 90 Coker 312 L4

y 70

t VipCot™ 0.80% 80 i l 60 VipCot™ L1 a 70 y t

t VipCot™ L4 r i

50 l o 60 a t M r 40 50 o %

30 M

40

20 % 30 10 20 0 10 Test 1 Test 2 Test 3 Test 4 Mean 0 Test 1Test 2 Test 3 Test 4 Fig. 1. Bioassay of H. virescens on lyophilized Coker312 and VipCotä leaf tissue. Neonate H. virescens larvae were exposed to 4 and 0.8% Fig. 3. Bioassay of L1 and L4 H. zea vs. Coker312 and VipCotä leaf preparations of lyophilized Coker 312 and VipCotä leaf tissue. 24 larvae tissue. L1 and L4 H. zea larvae were fed leaf tissue from Coker312 and were exposed to each treatment and mortality was assessed at three or four VipCotä cotton plants. 24 larvae of each instar were tested against each day intervals until the transgenic treatment mortality reached 100% or leaf type. Surviving larvae were provided with fresh tissue every two to there was no additional mortality in the transgenic treatment after an three days. Mortality was recorded at each tissue change until 100% approximately seven day period. Data are presented from four indepen- mortality was noted in the transgenic event. Data are presented for three dent repeats of the test, along with the mean data from all four tests. The independent repeats of the test, as well as the mean data across all three error bars indicate the standard deviations of the mean data. tests. The error bars indicated the standard deviations of the mean data. R.W. Kurtz et al. / Journal of Invertebrate Pathology 95 (2007) 227–230 229 allele to the PIP. In dose response assays to Vip3A and ground starting approximately 1 meter in either direction Cry1Ab, 4th instar H. zea were at least twenty five-fold from the plants containing larvae. Seven days later (four- more tolerant to the toxins than neonates (data not teen DAI), those plants containing larvae plus all plants shown). in the same row within 1 meter either direction were reas- Fig. 3 illustrates the results of bioassays of L1 and L4 H. sessed to observe any surviving larvae. At fourteen DAI, zea larvae on VipCotä leaf tissue samples. Data from three 20 H. virescens larvae and 14 H. zea larvae were observed independent experiments are presented, as well as the mean in the Coker312 (only a small section of row was examined data across all experiments. All three tests resulted in 100% for larvae) across all locations. In contrast, no larvae were mortality of both neonate and L4 H. zea larvae on Vip- found upon re-examination of VipCotä cotton plants Cotä leaf tissue, whereas both instars survived well on con- fourteen DAI. This indicates that the few small larvae that trol Coker312 tissue. This meets the criterion for were still alive on VipCotä cotton plants seven DAI did demonstrating a high dose for VipCotä against H. zea fol- not survive and that VipCotä cotton plants showed lowing US EPA Method 5. 100% mortality of H. virescens and H. zea fourteen DAI. This meets the criterion for demonstrating a high dose 2.3. Method 4: Artificial field infestations for VipCotä against H. virescens and H. zea following US EPA method 4. US EPA Method 4 involves artificially infesting field plots with laboratory strains of pest insects to determine 3. Cross-resistance studies if the PIP is expressed at a dose that kills greater than 99.9% of infested insects. VipCotä cotton was evaluated The IRM benefits associated with the cultivation of a in studies located at Leland, MS; Winnsboro, LA; and pyramided cotton trait expressing two insecticidal proteins Beasley, TX in 2006. Data combined across locations are can only be fully realized if cross-resistance between the in Table 1. Across all locations, approximately 7000 Vip- two component proteins does not exist. To understand Cotä plants were infested with H. virescens eggs and the potential for cross-resistance between Vip3A and Cry approximately 6000 VipCotä plants were infested with proteins, Syngenta undertook a number of investigative H. zea eggs. This resulted in approximately 275,000 and studies. First, Syngenta scientists determined the sequence 230,000 larvae, respectively, hatching on these plants. Sub- of the Vip3A gene. These studies show a complete lack of stantially smaller numbers of Coker312 plants were sequence homology between Vip3A and the Cry toxins infested in parallel, so that approximately 52,000 H. vires- and predict marked structural differences between them cens and 35,000 H. zea larvae hatched on approximately (Estruch et al., 1996). This strongly supports the conten- 750–900 plants. Infestations of the nontransgenic Coker312 tions that Vip3A is structurally distinct from all the Cry plots resulted in the successful establishment of both H. toxins and consequently that cross-resistance between virescens and H. zea populations at each location. A total Vip3A and any Cry toxin is highly unlikely. of 1092 H. virescens and 200 H. zea larvae were observed Next, Syngenta scientists and others established that the seven days after infestation (DAI) across all locations on manner of pore formation in the midgut epithelium result- the nontransgenic plants. These data confirm both that ing from Vip3A’s action, and the flux properties of those the infestation regime was effective and that the insects pores, are both different from those of channels formed infested were vigorous and healthy. However, despite from Cry1A toxin action and, by extension, perhaps from greater than seven times more plants being infested with other homologous Cry-type toxins (Estruch et al., 1996; approximately five to six times more insects only 46 H. Lee et al., 2003). Thus, Vip3A differs functionally from virescens and 26 H. zea larvae were observed on VipCotä the Cry toxins and has a low likelihood of cross-resistance cotton plants seven DAI across all locations. At each loca- to Cry1A. tion, larvae surviving on VipCotä cotton plants seven DAI Receptor competition binding studies and ligand bind- were very small, ranging from first to early second instars. ing studies with aminopeptidase-N and cadherin-like pro- Where surviving larvae were found on VipCotä cotton, the teins confirm that Vip3A does not recognise Cry1Ab or plants containing the larvae were tagged. To restrict larval Cry1Ac receptors in a number of lepidopteran species movement, plants in the same row were pulled out of the including the target pests, H. virescens and H. zea. Con-

Table 1 Summary of larval survival following artificial infestation of VipCotä and Coker312 plants with H. virescens or H. zea eggs. Insect Treatment Total larvae infested Number of plants assessed Insects observed 7 DAI Insects observed 14 DAI H. virescens Coker312 52,478 898 1092 (8244)a 20 VipCotä 275,584 6729 46 0 H. zea Coker312 35,652 768 200 (1606)a 14 VipCotä 230,164 5824 26 0 a Estimated number of larvae that would have been observed if an equal number of plants were sampled for Coker312 as for VipCotä. 230 R.W. Kurtz et al. / Journal of Invertebrate Pathology 95 (2007) 227–230 versely, Cry1A toxins do not recognize or interact with the opment to VipCotä, and the introduction of VipCotä Vip3A receptor (Lee et al., 2006). These observations should benefit overall resistance management for Bt strongly indicate that target site cross-resistance between expressing traits by offering an alternative choice for trans- the two VipCotä component proteins is highly unlikely. genic control of lepidopteran cotton pests. Hence, should a target site resistance occur, irrespective of the toxin to which it first developed, it would be most Acknowledgements unlikely to confer resistance to the second toxin in VipCotä. The authors thank our colleagues at Syngenta for devel- Finally, strains of H. virescens that have multiple mech- oping the data presented in this paper and Dr. Ryan Jack- anisms of resistance to Cry toxins, including those with son for sharing his data with us prior to publication. altered protease activity were bioassayed with Vip3A pro- tein. Processing of Vip3A is apparently unaffected by the References presence of this mechanism, since it overcomes all forms of resistance to the Cry toxins that are present in the resis- Caprio, M.A., 1998. Evaluating resistance management strategies for tant strains (Jackson et al., 2007). Accordingly, we believe multiple toxins in the presence of external refuges. J. Econ. Entomol. that should such a mechanism of resistance to Cry toxins 91, 1021–1031. arise in a field population, it is unlikely that the Vip3A com- Estruch, J.J., Warren, G.W., Mullins, M.A., Nye, G.J., Craig, J.A., ponent of VipCotä would be affected. In this context, it is Koziel, M.G., 1996. Vip3A, a novel Bacillus thuringiensis vegetative insecticidal protein with a wide sectrum of activities against relevant that Syngenta’s VipCotä is the only stacked insec- lepidopteran insects. Proc. Natl. Acad. Sci. USA 93, 5389–5394. ticidal cotton variety not expressing two Cry toxins. We Jackson, R.E., Marcus, M.A., Gould, F., Bradley, J.R., Van Duyn, J.W., believe this gives a significantly greater margin of safety 2007. Cross-resistance responses of Cry1Ac-selected Heliothis vires- against the possibility of cross-resistance between the two cens (Lepidoptera: Noctuidae) to the Bacillus thuringiensis protein components than all other stacked traits currently available. Vip3A. J. Econ. Entomol. 100 (1), 180–186. Lee, M.-K., Walters, F.S., Hart, H., Palekar, N., Chen, J.-S., 2003. The mode of action of the Bacillus thuringiensis vegetative insecticidal 4. Conclusions protein Vip3A differs from that of Cry1Ab (d-endotoxin. Appl. Environ. Microbiol. 69, 4648–4657. Herein, we have reviewed evidence based on two of the Lee, M.-K., Miles, P., Chen, J.-S., 2006. Brush border membrane binding US EPA methods that VipCotä expresses a high dose of properties of Bacillus thuringiensis Vip3A toxin to Heliothis virescens and Helicoverpa zea midguts. Biochem. Biophys. Res. Comm. 339, protein against H. virescens, and three methods that Vip- 1043–1047. Cotä expresses a high dose of protein against H. zea. O’Reilly, D., Mascarenhas, V., Green, M., Townley, A., Potts A,. Additionally, in O’Reilly et al. (2006), we have demon- McCaffery, A., 2006. Determination of the high dose status of cotton strated high dose for VipCotä component events against events expressing Cry1Ab. In: Proceedings of the 2006 Beltwide H. virescens, H. zea and Pectinophora gossypiella (pink Cotton Conference. National Cotton Council, Memphis, TN. O’Reilly, D., Walters, F., Palekar, N., Boyer, A., Kurtz, R., 2007. bollworm). Thus, we have robust demonstrations that Vip- Laboratory studies to determine the high dose status of VipCotä Cotä is high dose against these insect species. We have also cotton. In: Proceedings of the 2007 Beltwide Cotton Conference. summarized the available literature assessing the cross- National Cotton Council, Memphis, TN. resistance potential between Vip3A and Cry toxins. The Roush, R.T., 1997. Bt transgenic crops: just another pretty or a evidence from these studies indicates that Vip3A is struc- chance for a new start in resistance management? Pestic. Sci. 51, 328– 334. turally and functionally distinct from Cry toxins and that Roush, R.T., 1998. Two-toxin strategies for management of insecticidal there is a complete lack of cross-resistance between Vip3A transgenic crops: Can pyramiding succeed where mixtures and the Cry proteins tested. Consequently, Vip3A and have not? Phil. Trans. R. Soc. Lond. B 353, 1777–1786. Cry1Ab are excellent partners as a pyramided PIP, and US EPA, 1998. Bacillus thuringiensis (B.t.) plant- and resistance combining the two proteins in the same plant should have management. United States Environmental Protection Agency, Report number EPA 738-F-98-001. a beneficial impact on IRM for the VipCotä trait. Taking US EPA, 2001. Bt plant-incorporated protectants. October 15, 2001 all of this work together, a high dose plus refuge-based Biopesticides Registration Action Document. http://www.epa.gov/ IRM strategy will be effective in delaying resistance devel- pesticides/biopesticides.