Identification of a New cry1I-Type Gene as a Candidate for Gene Pyramiding in Corn To Control Ostrinia Species Larvae

Abstract

Pyramiding of diverse cry toxin genes from Bacillus thuringiensis with different modes of action is a desirable strategy to delay the evolution of resistance in the European corn borer (Ostrinia nubilalis). Considering the dependency of susceptibility to Cry toxins on toxin binding to receptors in the midgut of target pests, a diverse mode of action is commonly defined as recognition of unique binding sites in the target insect. In this study, we present a novel cry1Ie toxin gene (cry1Ie2) as a candidate for pyramiding with Cry1Ab or Cry1Fa in corn to control Ostrinia species larvae. The new toxin gene encodes an 81-kDa protein that is processed to a protease-resistant core form of approximately 55 kDa by trypsin digestion. The purified protoxin displayed high toxicity to Ostrinia furnacalis and O. nubilalis larvae but low to no activity against Spodoptera or heliothine species or the coleopteran Tenebrio molitor. Results of binding assays with ¹²⁵I-labeled Cry1Ab toxin and brush border membrane vesicles from O. nubilalis larvae demonstrated that Cry1Ie2 does not recognize the Cry1Ab binding sites in that insect. Reciprocal competition binding assays with biotin-labeled Cry1Ie2 confirmed the lack of shared sites with Cry1Ab or Cry1Fa in O. nubilalis brush border membrane vesicles. These data support Cry1Ie2 as a good candidate for pyramiding with Cry1Ab or Cry1Fa in corn to increase the control of O. nubilalis and reduce the risk of resistance evolution.

INTRODUCTION

Insecticidal crystal (Cry) proteins produced by the bacterium Bacillus thuringiensis during sporulation have been widely used as part of spray products or expressed in transgenic crops to control devastating insect pests (1). The mode of action of Cry toxins includes solubilization and activation in the larval midgut fluids, followed by binding to protein receptors on the midgut epithelium, resulting in toxin insertion into the membrane and the formation of a pore that results in enterocyte death by osmotic shock (2). Alterations in any of these steps could potentially lead to insect resistance, yet high levels of resistance are always linked to alterations in toxin receptor genes (3).

Pyramiding of multiple B. thuringiensis toxin genes recognizing distinct binding sites in the midgut of the target pest has been demonstrated to delay the evolution of resistance in transgenic plants (4). On the basis of this strategy, transgenic corn producing combinations of the Cry1Ab or Cry1Fa protein or the chimeric Cry1A.105 protein (domains I and II from Cry1Ab and domain III of Cry1Fa) has been commercialized for control of European corn borer (Ostrinia nubilalis) larvae. However, recent reports support the idea that Cry1Ab, Cry1Fa, and Cry1A.105 have binding sites in the O. nubilalis midgut in common (5). Moreover, the genetic potential to develop resistance to Cry1Ab and Cry1Fa has been demonstrated for strains of O. nubilalis through laboratory selection (6–8). These observations, together with the high adoption of B. thuringiensis-treated corn (Bt corn), currently comprising 80% of the corn planted in the United States (http://www.ers.usda.gov/data-products/adoption-of-genetically-engineered-crops-in-the-us/recent-trends-in-ge-adoption.aspx), underlines the high probability of the development of resistance in O. nubilalis. Identification of new Cry toxins amenable to pyramiding with the Cry1A or Cry1Fa toxin in Bt corn is crucial to the development of novel plants that delay resistance evolution in O. nubilalis.

Among the cry genes identified to date (http://www.lifesci.sussex.ac.uk/home/Neil_Crickmore/Bt/), the cry1I family (formerly cryV genes) includes genes with unique properties. For instance, they are usually silent genes in B. thuringiensis strains but can be expressed in Escherichia coli cultures as a protoxin form of approximately 81 kDa (9), a unique molecular mass among Cry1 toxins. Moreover, different Cry1I-type proteins have been identified as active against both lepidopteran and coleopteran pests (10, 11). Importantly, currently available data support the idea that insects resistant to Cry1A toxins do not display cross-resistance to Cry1I toxins (9, 11). The protein encoded by the first cry1Ie gene described (Cry1Ie1) was reported as toxic to larvae of Plutella xylostella, Ostrinia furnacalis, and Leguminivora glycinivorella (12). Transgenic maize expressing Cry1Ie1 was shown to effectively control O. furnacalis in field trials and displayed toxicity to Cry1Ac-resistant Helicoverpa armigera (13). Furthermore, no cross-resistance to Cry1Ie was detected in a Cry1Ab-resistant strain of O. furnacalis (14) or a Cry1Ac-resistant strain of P. xylostella (15). Expression of the Cry1Ac and Cry1Ie1 toxins in transgenic tobacco produced greater activity against susceptible or Cry1Ac-resistant H. armigera than in plants producing a single toxin (16). Qualitative binding competition assays demonstrated that Cry1Ie1 and Cry1Ac have no binding sites in common in midgut brush border membrane vesicles (BBMV) from O. furnacalis (9). Taken together, these data support Cry1Ie1 as a good candidate for pyramiding with Cry1Ac. In contrast, no data are available on the sharing of binding sites by Cry1Ie and the Cry1Ab and Cry1Fa toxins, which are currently used in transgenic corn to control O. nubilalis.

In the present study, a new cry1Ie gene (cry1Ie2) was cloned from B. thuringiensis strain T03B001 and expressed successfully in E. coli. Insect bioassays with the purified Cry1Ie2 protoxin confirmed that the toxin is highly active against Ostrinia species larvae. Results of binding competition assays supported Cry1Ie2 as an optimal candidate for pyramiding with Cry1Ab or Cry1Fa in corn to control O. nubilalis larvae and delay the evolution of resistance.

MATERIALS AND METHODS

Cloning of the cry1Ie2 gene.

The full-length cry1Ie2 gene was cloned from genomic DNA extracted from B. thuringiensis strain T03B001 from the Bacillus Genetic Stock Center (BGSC; Columbus, OH) as described elsewhere (17). The primers used for amplification by PCR were 1Ie5 (5′-ATGAAATCGAAGAATCAAGATATGTATC-3′) and 1Ie3 (5′-CATGTCACGCTCAATATTGAG-3′). Amplification was performed with Phusion DNA polymerase (NEB Corporation, China) and a PTC-100 Peltier thermal cycler (MJ Research) as follows: 30 cycles of denaturation at 94°C for 1 min, annealing at 50°C for 1 min, and extension at 72°C for 2.5 min, followed by an additional extension at 72°C for 10 min. After amplification, the full-length PCR products were cloned into the pMD19-T vector for sequencing with an automated DNA sequencer (ABI-3730XL) and VectorNTI Suite 9 (Invitrogen, Carlsbad, CA) for sequence analysis. The cloned cry1Ie2 gene was then subcloned into the Ecl136II site of the pEB vector (18), and the resulting construct was transformed into E. coli JM109 cells. After sequencing, successful transformants were used to produce a plasmid, which was purified with a plasmid extraction kit (Axygen Biotechnology Corporation, China) and then transformed into E. coli Rosetta(DE3) (TaKaRa Corp., China) for expression.

Cry1Ie, Cry1Ab, and Cry1Fa toxin purification.

Expression of cry1Ie2 in 500-ml LB cultures of E. coli was induced with 0.4 mM isopropyl-β-d-thiogalactopyranoside (IPTG) at 20°C for approximately 15 h, and then cells were collected by centrifugation at 5,000 × g for 10 min and the cell pellet was resuspended in 25 ml of solubilization buffer (50 mM Na₂CO₃ with 0.1% 2-mercaptoethanol and 0.1 M NaCl, pH 10.0). Cells were then lysed by ultrasonication (Ningbo Scientz Biotechnology Co., Ltd., China) for 6 min (75% power, 3-s pulse on, 5-s pulse off) and then solubilized for 15 h at 30°C with shaking at 150 rpm. The liberated inclusion bodies were separated from the supernatant by centrifugation at 15,000 × g for 30 min at 4°C. Both soluble and insoluble protein fractions were examined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as previously described (19). The Cry1Ie2 protoxin (∼81 kDa) was greatly enriched in the soluble fraction and was purified by affinity chromatography in a 1-ml HisTrap HP column (GE Healthcare) preequilibrated with binding buffer (50 mM imidazole, 20 mM Tris-HCl, 500 mM NaCl, pH 8.5). The column flowthrough was collected and reloaded to improve recovery of the protein. Nonspecifically adsorbed proteins were removed by washing with 2 column volumes of binding buffer. Proteins bound to the column were eluted with a linear gradient of 0 to 20% elution buffer (500 mM imidazole, 20 mM Tris-HCl, 500 mM NaCl, pH 8.5), followed by a step gradient of 100% elution buffer while monitoring for elution on the basis of absorbance at 280 nm. The fractions included in absorbance peaks were collected and analyzed by 10% SDS-PAGE. Imidazole was removed from the fractions containing Cry1Ie2 protoxin (as detected by SDS-PAGE) via dialysis in 50 mM Na₂CO₃ buffer, pH 9.8. Protein concentration was determined with the Qubit Protein Assay kit in a Qubit fluorometer (Life Technologies) by following the manufacturer's instructions, and then protein was stored at −80°C until used (less than 3 months) for insect bioassays.

The Cry1Ie2 protoxin was activated by treatment with trypsin (tosylsulfonyl phenylalanyl chloromethyl ketone [TPCK] treated, 1:5 [wt/wt]) in 50 mM Na₂CO₃ (pH 10) at 37°C for 1 h. The trypsin-digested products were loaded onto a 1-ml HiTrap Q HP column (GE Healthcare) that had been preequilibrated with buffer A (50 mM Na₂CO₃, pH 10.2). The column was then washed with 2 to 3 column volumes of buffer A, and activated toxin was eluted with a linear gradient of 0 to 1 M NaCl in buffer A. The fractions included in absorbance peaks were collected and analyzed by 10% SDS-PAGE. Fractions containing Cry1Ie2 toxin (∼55 kDa) were pooled, and the toxin was stored at −80°C until used for labeling and binding assays.

An E. coli strain harboring the cry1Ab gene from the B. thuringiensis subsp. kurstaki HD-1 strain (20) was obtained from the BGSC (Columbus, OH). Expression, activation, and purification of Cry1Ab were performed as described for Cry1Ie, except for the use of histidine tag-based purification with HisTrap HP columns.

Recombinant B. thuringiensis acrystalliferous mutant strain HD-73^- harboring the cry1Fa gene and an E. coli Rosetta(DE3) strain harboring the cry1Ie1 gene from the B. thuringiensis Btc007 strain (12) were from our laboratory at the Institute of Plant Protection, Chinese Academy of Agricultural Sciences (Beijing, China). Expression, activation, and purification of Cry1Fa were performed as described previously (21), and Cry1Ie1 was purified as described above for Cry1Ie2.

Insect bioassays.

The insecticidal activity of the Cry1Ie2 protoxin was tested against neonates of O. furnacalis, O. nubilalis, H. armigera, Spodoptera exigua, Spodoptera frugiperda, Heliothis virescens, Helicoverpa zea, and T. molitor. Larvae of O. furnacalis, H. armigera, and S. exigua were kindly supplied by the Cotton and Corn Insect Pest laboratories at the Institute of Plant Protection, Chinese Academy of Agricultural Sciences (Beijing, China). Eggs of O. nubilalis, S. frugiperda, H. virescens, and H. zea were purchased from Benzon Research (Carlisle, PA). Larvae of T. molitor were obtained from a laboratory colony maintained by the Department of Entomology and Plant Pathology at the University of Tennessee (Knoxville, TN). Bioassays were replicated three times, with each toxin concentration tested with at least 32 insects, and deaths were determined after 7 days at 27°C with a 16-h light and 8-h dark cycle and 60 to 80% relative humidity. Deaths were corrected for natural deaths (deaths in controls) by using Abbott's formula (22). The POLO-PC software package (23) was used for probit analysis to estimate 50% lethal concentrations (LC₅₀s).

For O. furnacalis, O. nubilalis, H. armigera, and S. exigua bioassays, neonate larvae were exposed to a meridic diet incorporating seven dilutions of Cry proteins as described elsewhere (24). Bioassays were performed with 128-well trays, with each well containing a plug of approximately 400 mg of diet. Insecticidal activity against H. virescens, S. frugiperda, and H. zea larvae was tested by surface contamination with 128-well bioassay trays (C-D International, Pitman, NJ) as previously described (25, 26). A volume (75 μl) of test solution was added to the top of the solid meridic diet (General Purpose Lepidoptera diet; BioServ) previously poured into each well, spread evenly on the surface, and allowed to dry in a laminar-flow hood. After diet drying, one neonate was transferred to each well with a fine camel's hair paintbrush and then the well was sealed with an adhesive bioassay tray lid. Bioassays with T. molitor neonates were performed as described elsewhere (27).

Preparation of BBMV.

Midguts were dissected from fifth-instar O. nubilalis larvae, and BBMV were prepared by a differential centrifugation method (28), with minor modifications (29). The protein concentration in isolated BBMV was determined by fluorometry as described above for protoxin quantitation, and BBMV proteins were stored in aliquots at −80°C until used (less than 2 months). Purity of BBMV preparations was determined by estimating the enrichment of aminopeptidase N (APN) specific activity in the BBMV compared to that in initial midgut tissue homogenates, as described elsewhere (30). Representative APN activity enrichment in the final BBMV preparations was 3- to 5-fold compared to initial midgut homogenates.

Toxin labeling.

Trypsin-activated and purified Cry1Ab toxin (20 μg) was labeled with 0.5 mCi ¹²⁵I (PerkinElmer) by using chloramine T as described elsewhere (31). The specific activity of labeled Cry1Ab was 3.5 μCi/μg on the basis of the initial toxin input.

Purified Cry1Ie2 toxin (1 mg) was labeled with biotin by incubation with 10 nM EZ-Link NHS-LC-Biotin (Thermo Scientific) in phosphate-buffered saline (PBS) buffer (137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 1.8 mM KH₂PO₄, pH 7.4) on ice for 2 h. Free biotin was removed by dialysis overnight in a total of 4 liters of 20 mM Na₂CO₃ (pH 9.8)–150 mM NaCl at 4°C. Purified biotinylated Cry1Ie2 was quantified by fluorometry as described above for protoxin quantitation and stored in aliquots at −80°C until used for binding assays.

Binding assays with ¹²⁵I-labeled Cry1Ab toxin.

Saturation binding assays to test specific toxin binding and to specify an optimal concentration of ¹²⁵I-labeled Cry1Ab for competition binding assays were performed with a constant concentration of O. nubilalis BBMV proteins (0.2 mg/ml) and increasing concentrations of ¹²⁵I-labeled Cry1Ab in a final volume of 0.1 ml of binding buffer (PBS plus 0.1% bovine serum albumin [BSA]). Binding reaction mixtures were incubated for 1 h at room temperature, and then toxin bound to BBMV was separated from free toxin by centrifugation (13,500 × g for 10 min). The BBMV pellets were resuspended in 1 ml of ice-cold binding buffer and subjected to one more centrifugation step. The radioactivity remaining in the final BBMV pellets was measured in a Wizard² gamma counter (PerkinElmer) and considered total binding. Nonspecific binding was determined by including a 100-fold excess of unlabeled Cry1Ab toxin in the binding reactions. Specific binding was calculated by subtracting nonspecific binding from total binding.

Homologous and heterologous binding competition experiments were performed by incubating 20 μg of O. nubilalis BBMV with 1 nM ¹²⁵I-labeled Cry1Ab toxin and increasing concentrations of unlabeled competitor (Cry1Ab or Cry1Ie2 toxin) in a final volume of 0.1 ml of binding buffer for 1 h at room temperature. Binding reactions were stopped by centrifugation, and samples were processed as described for saturation binding assays.

Each binding experiment was conducted in duplicate and replicated at least twice. Binding data were analyzed and plotted with SigmaPlot v.11.0 software (Systat Software). Fitting of the specific binding data by nonlinear regression determined a one-site model as the best fit to the data, which was used to determine the apparent Cry1Ab dissociation constant (K_d) and concentration of binding sites (B_max).

Binding assays with biotinylated Cry1Ie2 toxin.

Twenty micrograms of BBMV proteins from O. nubilalis was incubated with 0.1 μg of biotinylated Cry1Ie2 in the presence or absence of a 3- to 500-fold excess of unlabeled toxins in a final volume of 100 μl of binding buffer (PBS plus 0.1% BSA) for 1 h at room temperature. Binding reactions were stopped by centrifugation for 15 min at 13,500 × g, after which BBMV and bound toxin pellets were washed with 1 ml of ice-cold binding buffer. After centrifugation, pellets were solubilized in 15 μl of sample buffer (32), heat denatured for 10 min, and loaded onto Criterion TGX Precast Gels (Bio-Rad). After electrophoretic separation, proteins were electrotransferred to Hybond low-fluorescence polyvinylidene difluoride (PVDF) blotting filters (GE Healthcare Life Sciences) as described elsewhere (26). After blocking for 1 h at room temperature in PBS buffer containing 5% dried milk powder and 0.1% Tween 20, filters were probed with streptavidin-Alexa Fluor 594 conjugate (Invitrogen) at a 1:5,000 dilution for 3 h at room temperature. Filters were washed with washing buffer for 1 h (10 min per wash) and air dried on filter paper. The dry filters were scanned for red fluorescence (633 nm) in a Typhoon Trio scanner (GE Healthcare). Each competition experiment was replicated at least twice. Densitometric analysis of the fluorescent bands detected in the blot was performed with ImageJ software (33) to quantify binding. Data were analyzed for significance and plotted with SigmaPlot v.11.0 software (Systat Software).

Nucleotide sequence accession number.

The cry1Ie sequence described here has been submitted to GenBank and assigned accession number HM439636.

RESULTS

Cloning and sequencing of the cry1Ie2 gene.

A cry1Ie gene was identified by genomic sequencing of B. thuringiensis strain T03B001 (BGSC no. 4AO1), which was originally reported as being toxic to larvae of Bombyx mori, Hyphantria cunea, and Clostera anastomosis (34). The 2,157-bp full-length cry1Ie gene was successfully cloned and sequenced (GenBank accession number HM439636). It encoded a polypeptide of 719 amino acids with a predicted molecular mass of 81.4 kDa and 96% sequence identity to the Cry1Ie1 protein (GenBank accession number AAG43526). The new toxin sequence was submitted to the B. thuringiensis Delta-Endotoxin Nomenclature Committee (http://www.lifesci.sussex.ac.uk/home/Neil_Crickmore/Bt/), which assigned the name Cry1Ie2 to the new protein.

Expression of the cry1Ie2 gene.

The cry1Ie2 gene was successfully expressed in E. coli Rosetta(DE3) as an 81-kDa protein, which was detected in both inclusion bodies and the soluble fraction (Fig. 1A). As is the case with other Cry proteins, the insoluble Cry1Ie2 protein was soluble in alkaline buffer in the presence of a reducing reagent. Both Cry1Ie2 and the closely related Cry1Ie1 protoxin contain a trypsin splice site at arginine 153, which resulted in a protease-resistant Cry1Ie2 core form of approximately 55 kDa (Fig. 1B) after processing with bovine trypsin, as observed for Cry1Ie1 (35). A much less abundant 66-kDa fragment, probably representing the partially processed Cry1Ie2 protoxin, was initially observed among the products of trypsin digestion but disappeared upon longer trypsin incubations (data not shown), as reported for Cry1Ie1 (35).

FIG 1.

Analysis of Cry1Ie2 toxin production, purification, and labeling. (A) SDS-PAGE analysis of Cry1Ie2 protoxin production in E. coli cultures. The samples shown are the soluble and insoluble fractions of E. coli cultures transformed with the empty expression construct (pEB) or with the construct including the cry1Ie2 gene (Cry1Ie2). The arrow indicates the expected molecular size of the Cry1Ie2 protoxin form. (B) Purified protease-resistant Cry1Ie2 toxin core generated after digestion with trypsin (Cry1Ie2 toxin) was labeled with biotin (Biotin-Cry1Ie2), and labeling was checked by Western blotting (Blot), probing with a streptavidin-horseradish peroxidase conjugate, and development by enhanced chemiluminescence. The values to the left of each panel are molecular sizes in kilodaltons. MWM, molecular mass markers.

Insect bioassays.

The relative insecticidal activity of the purified Cry1Ie2 protoxin was tested and compared to that of the Cry1Ie1 protoxin. Activity against seven lepidopteran and one coleopteran species was tested. The highest relative insecticidal activity of both Cry1Ie1 and Cry1Ie2 was detected against larvae of O. furnacalis and O. nubilalis (Table 1). In contrast, no toxicity to larvae of S. exigua, H. zea, or T. molitor was detected (Table 1). While no H. armigera, H. virescens, or S. frugiperda deaths were detected, clear growth inhibition was observed (data not shown), suggesting low toxicity.

TABLE 1.

Bioassay results for CryIe1 and Cry1Ie2 toxins against larvae of diverse lepidopteran and coleopteran pests^a

Assay and insect	LC50 (95% fiducial limits)
	Cry1Ie2	Cry1Ie1
Diet incorporation bioassay
O. furnacalis	0.51 (0.22–1.04)	NTb
O. nubilalis	18.49 (11.15–27.36)	18.58 (13.66–23.71)
H. armigera	>100	>100
S. exigua	>100	>100
Surface contamination bioassay
H. virescens	>7,500	>7,500
H. zea	>7,500	>7,500
S. frugiperda	>7,500	>7,500
T. molitor	>384c	>384c

^aToxicity values, expressed as LC₅₀s in μg/ml for diet incorporation bioassays and in ng/cm² for surface contamination bioassays, unless specified otherwise, are reported only for active toxins. The concentrations of other toxins represent the highest toxin concentrations tested that resulted in a <50% mortality rate.

^bNT, not tested.

^cValue is in μg/disk.

Binding assays with ¹²⁵I-labeled Cry1Ab toxin.

In saturation binding experiments, binding of ¹²⁵I-labeled Cry1Ab to BBMV proteins from O. nubilalis larvae was specific and saturable (Fig. 2A). Specific Cry1Ab binding appeared to saturate after the 3 nM ligand concentration. Fitting of binding data to a one-site binding model equation resulted in an apparent K_d of 2.55 ± 0.68 nM and a B_max of 2.28 ± 0.29 pmol/mg BBMV for ¹²⁵I-labeled Cry1Ab.

FIG 2.

Saturation (A) and competition (B) binding experiments with BBMV proteins (20 μg) from O. nubilalis and ¹²⁵I-labeled Cry1Ab toxin. (A) Specific binding of increasing ¹²⁵I-labeled Cry1Ab concentrations to BBMV proteins was calculated by subtracting nonspecific from total binding in the absence of unlabeled competitor. Data shown are the means and standard errors from at least two independent experiments performed in duplicate for each ¹²⁵I-labeled Cry1Ab concentration. The curve shown was fitted to the specific binding data by using a one-site binding model as the best-fitting model. (B) Curves represent total binding of ¹²⁵I-labeled Cry1Ab (1 nM) to BBMV in the presence of increasing concentrations of the unlabeled Cry1Ab or Cry1Ie2 toxin competitor, as indicated. Each competition experiment was replicated at least twice, and the error bars represent the standard error of the mean.

Competition of ¹²⁵I-labeled Cry1Ab binding by unlabeled Cry1Ab demonstrated specific binding, as observed in saturation assays (Fig. 2B). In contrast, unlabeled Cry1Ie2 did not displace ¹²⁵I-labeled Cry1Ab binding to O. nubilalis BBMV, even at the highest Cry1Ie2 concentration tested (Fig. 2B), supporting the idea that Cry1Ie2 does not recognize Cry1Ab binding sites in O. nubilalis BBMV.

Binding competition assays with biotinylated Cry1Ie2 toxin.

Preliminary binding assays with radiolabeled Cry1Ie2 toxin determined that iodination greatly affects the binding of the toxin to BBMV (data not shown). As an alternative, we used biotinylated Cry1Ie2 toxin (Fig. 1B) to perform binding competition assays. Binding of biotinylated-Cry1Ie2 toxin to O. nubilalis BBMV was displaced by increasing concentrations of unlabeled homologous (Cry1Ie2) competitor (Fig. 3A), supporting the specific binding of this toxin. In contrast, unlabeled Cry1Ab or Cry1Fa toxin did not reduce the binding of biotinylated Cry1Ie2 to O. nubilalis BBMV (Fig. 3A). Densitometry of binding competition assay results (Fig. 3B) supported the idea that neither the Cry1Ab nor the Cry1Fa toxin has the same binding sites as the Cry1Ie2 toxin in O. nubilalis BBMV.

FIG 3.

Competition binding assay with the biotinylated Cry1Ie2 (18 nM) and BBMV (20 μg) proteins from O. nubilalis. (A) Detection of bound biotinylated Cry1Ie2 by Western blotting in competition binding assays with increasing concentrations of unlabeled Cry1Ie2, Cry1Ab, or Cry1Fa competitor, as indicated. The competitor concentrations shown are expressed as 0- to 500-fold excesses with respect to the biotinylated Cry1Ie2 input in all reaction mixtures (18 nM). After binding reactions, proteins were separated by electrophoresis and transferred to low-fluorescence PVDF blotting filters, and bound biotinylated Cry1Ie2 was detected by probing filters with streptavidin-Alexa Fluor 594 conjugate and scanning for red fluorescence. (B) Densitometric analysis of the bands in panel A with MacBiophotonics ImageJ software (Systat Software) to quantify binding. Shown are the mean total binding and corresponding standard errors calculated from at least two independent experiments. Total competitor concentrations (nM), rather than fold excesses, as in panel A, are shown in the graph.

DISCUSSION

Transgenic maize producing Cry proteins from B. thuringiensis has successfully controlled O. nubilalis populations (36, 37). While no field-evolved practical resistance in O. nubilalis has been reported to date, data on laboratory selection support the ability of this insect to develop resistance to Cry toxins currently produced by Bt corn, including Cry1Ab (7, 8, 38) and Cry1Fa (39). Pyramiding of the cry1Ab and cry1Fa toxin genes in Bt corn to delay the evolution of resistance in O. nubilalis was initially supported by binding and bioassay data suggesting that these toxins have unique sites in O. nubilalis BBMV (40) and that levels of cross-resistance to these toxins are low (7, 39). However, recent data support the idea that most of the Cry1Ab and Cry1Fa binding sites in O. nubilalis BBMV are the same (5), implying a high risk of the development of resistance to Cry1Ab-Cry1Fa pyramided Bt corn by alteration of shared binding sites. Consequently, new B. thuringiensis toxins with binding sites different from those recognized by Cry1Ab or Cry1Fa in O. nubilalis BBMV are needed to develop effective pyramided corn varieties.

In this study, we characterize a new Cry1Ie toxin displaying high relative activity against the European and Asian corn borers. The levels of Cry1Ie2 toxicity to O. nubilalis neonates were comparable to those reported for the Cry1Ab (41, 42) and Cry1Fa (5, 40) toxins. This similar toxicity is also indirectly supported by reports of a lack of significant differences (based on overlapping confidence intervals) in Cry1Ie1, Cry1Ab, and Cry1Fa toxicity to O. furnacalis (14) and the same toxicity to Ostrinia species larvae observed for the Cry1Ie1 and Cry1Ie2 toxins in the present study. The new Cry1Ie2 toxin was expressed in E. coli cultures as a protoxin of 81 kDa, a molecular mass uniquely common to other Cry1I toxins (11, 12). As in other Cry1I toxins (35), a protease-resistant core form of 55 kDa was observed. Comparison of the predicted secondary structure of Cry1Ie1 and Cry1Aa indicated that the N terminus of the Cry1Ie1 core form is located at the beginning of helix α-4 in domain I of the toxin (35). Proteolytic cleavage of helix α-1 in domain I of Cry1Ab was reported to be important for toxin oligomerization, and modified toxins lacking this helix formed oligomers in solution independently of interactions with receptors (43). This observation may explain the detection of oligomeric Cry1Ie1 (35) and Cry1Ie2 (data not shown) toxin forms in solution in the absence of any receptor preparation. However, previous reports support the idea that Cry1Ie1 oligomers are inactive against lepidopteran larvae (35), in contrast to the need for oligomerization for insecticidal activity in Cry1A toxins (44). Further research should determine if alternative Cry1Ie toxin oligomers conducive to toxicity are formed after interaction with toxin receptors, as well as the role of domain I differences between the Cry1A and Cry1Ie toxins in the formation of these oligomeric structures.

Although Cry1Ie2 differs from Cry1Ie1 by two amino acids in domain I, two amino acids in domain II, and four amino acids in domain III, we did not detect significant differences between the activity spectra of these toxins. Thus, both Cry1Ie1 and Cry1Ie2 were highly active against Ostrinia species but were inactive against or showed only growth inhibitory effects on other pests of corn from Asia and North America, including H. armigera, S. exigua, H. virescens, H. zea, and S. frugiperda. These observations are in contrast to a previous report on Cry1Ia7 toxin activity against noctuid larvae (11) but in agreement with bioassay data on Cry1Ie toxins (12). High Cry1Ie activity against Ostrinia species explains why Cry1Ie1-expressing maize effectively suppressed O. furnacalis in field trials (14). Taken together, our bioassay data support a restricted range of Cry1Ie2 activity against lepidoptera yet high activity against key Ostrinia pest species.

Previous reports demonstrated that Cry1I toxins are also active against larvae of the Colorado potato beetle (Leptinotarsa decemlineata) (10, 11) or corn rootworm (Diabrotica virgifera virgifera) (45). The lack of Cry1Ie2 toxicity to T. molitor larvae in our bioassays is in agreement with previous reports about this insect and Cry1Ia toxin (46) or Cry1Ie1 and elm leaf beetle (Pyrrhalta aenescens) larvae (12) and support the idea that the Cry1I toxin activity reported so far in Coleoptera is limited to some members of the Chrysomelidae family.

Data from competition binding assays using radiolabeled Cry1Ab in this study demonstrate that Cry1Ie2 does not recognize Cry1Ab binding sites in O. nubilalis BBMV. While we attempted to radioiodinate Cry1Ie2 to perform quantitative binding assays, we were unable to obtain an iodinated toxin preparation capable of binding specifically to lepidopteran BBMV. Similar observations were reported for Cry1Fa (21), yet binding was detected when that toxin was labeled with lower specific activity (47). We were unsuccessful, after attempting diverse labeling conditions and methods, including labeling with lower specific activity, to detect levels of specific binding amenable to the performance of competition binding assays (data not shown). This observation may suggest that tyrosine residues may be critical to Cry1Ie2 binding to the BBMV proteins. As an alternative, binding assays with biotinylated Cry1Ie2 demonstrated specific Cry1Ie2 binding to O. nubilalis BBMV sites and that these sites were not recognized by Cry1Ab or Cry1Fa. Since domain II in Cry toxins is the main determinant of their specificity (2), the lack of shared binding sites is also supported by low sequence similarity (<50%) between domains II of the Cry1Ie2 and Cry1Ab or Cry1Fa toxins. On the basis of its high identity with Cry1Ie2 (>90%), we expect that Cry1Ie1 would also not have receptors in common with the Cry1Ab toxin, which would explain the lack of cross-resistance to Cry1Ie1 in a Cry1Ab-resistant strain of O. furnacalis (14) and Cry1Ac-resistant strains of H. armigera (16) and Plutella xylostella (15). A lack of shared binding sites between another Cry1I toxin (Cry1Ia7) and Cry1Ab was previously reported qualitatively for BBMV from Lobesia botrana and Earias insulana (11). While not tested empirically and based on previously described similarities in shared binding sites among Ostrinia species (48), a lack of Cry1Ie2 and Cry1Ab or Cry1Fa binding sites in O. furnacalis BBMV would be expected. High activity, lack of shared binding sites, and reports of no cross-resistance among the Cry1A and Cry1Ie toxins support Cry1Ie2 as an excellent candidate for gene pyramiding with the Cry1A or Cry1Fa toxin in Bt corn for Ostrinia species control. Current efforts are focused on the identification of Cry1Ie-specific receptors in Ostrinia species larvae.