Mode of Action of Bipyramidal delta-Endotoxin of Bacillus thuringiensis subsp. kurstaki HD-1

The mode of action of the toxic fragment (P-59) derived from bipyramidal-shaped delta-endotoxin of Bacillus thuringiensis subsp. kurstaki HD-1 on the silkworm Bombyx mori was investigated. An enzyme-linked immunosorbent assay showed that there was no translocation of P-59 from the gut lumen to the hemocoel. When membrane vesicles prepared from silkworm midgut were incubated with P-59, normally smooth surface of vesicles became rough, and patch formation was observed on the surface. Vesicles treated with P-59 tended to agglutinate. The vesicle-denaturing activity of a 130,000-dalton subunit protein of bipyramidal toxin was enhanced by treatment with a gut juice protease of the silkworm. P-59 did not cause any uncoupling effect on mitochondria of the silkworm midgut. These results suggest that the attacking site of this toxin is not the mitochondrion but the cell membrane of the susceptible cell.     

Asporogenous Bacillus thuringiensis Mutant Producing High Yields of δ-Endotoxin

An asporogenous Bacillus thuringiensis subsp. kurstaki strain IK mutant, strain 290-1, which produced high yields of δ-endotoxin, was obtained by ethyl methane sulfonate treatment of a spore suspension. The mutant strain produced about the same amount of δ-endotoxin as that produced by the parent strain, but 10¹⁰ to 10¹¹ cells did not form any detectable dormant spores.     

Role of Proteolysis in Determining Potency of Bacillus thuringiensis Cry1Ac δ-Endotoxin

Bacillus thuringiensis protein δ-endotoxins are toxic to a variety of different insect species. Larvicidal potency depends on the completion of a number of steps in the mode of action of the toxin. Here, we investigated the role of proteolytic processing in determining the potency of the B. thuringiensis Cry1Ac δ-endotoxin towards Pieris brassicae (family: Pieridae) and Mamestra brassicae (family: Noctuidae). In bioassays, Cry1Ac was over 2,000 times more active against P. brassicae than against M. brassicae larvae. Using gut juice purified from both insects, we processed Cry1Ac to soluble forms that had the same N terminus and the same apparent molecular weight. However, extended proteolysis of Cry1Ac in vitro with proteases from both insects resulted in the formation of an insoluble aggregate. With proteases from P. brassicae, the Cry1Ac-susceptible insect, Cry1Ac was processed to an insoluble product with a molecular mass of ~56 kDa, whereas proteases from M. brassicae, the non-susceptible insect, generated products with molecular masses of ~58, ~40, and ~20 kDa. N-terminal sequencing of the insoluble products revealed that both insects cleaved Cry1Ac within domain I, but M. brassicae proteases also cleaved the toxin at Arg423 in domain II. A similar pattern of processing was observed in vivo. When Arg423 was replaced with Gln or Ser, the resulting mutant toxins resisted degradation by M. brassicae proteases. However, this mutation had little effect on toxicity to M. brassicae. Differential processing of membrane-bound Cry1Ac was also observed in qualitative binding experiments performed with brush border membrane vesicles from the two insects and in midguts isolated from toxin-treated insects.     

Production of Chymotrypsin-Resistant Bacillus thuringiensis Cry2Aa1 δ-Endotoxin by Protein Engineering

Cleavage of the Cry2Aa1 protoxin (molecular mass, 63 kDa) from Bacillus thuringiensis by midgut juice of gypsy moth (Lymantria dispar) larvae resulted in two major protein fragments: a 58-kDa fragment which was highly toxic to the insect and a 49-kDa fragment which was not toxic. In the midgut juice, the protoxin was processed into a 58-kDa toxin within 1 min, but after digestion for 1 h, the 58-kDa fragment was further cleaved within domain I, resulting in the protease-resistant 49-kDa fragment. Both the 58-kDa and nontoxic 49-kDa fragments were also found in vivo when ¹²⁵I-labeled toxin was fed to the insects. N-terminal sequencing revealed that the protease cleavage sites are at the C termini of Tyr49 and Leu144 for the active fragment and the smaller fragment, respectively. To prevent the production of the nontoxic fragment during midgut processing, five mutant proteins were constructed by replacing Leu144 of the toxin with Asp (L144D), Ala (L144A), Gly (L144G), His (L144H), or Val (L144V) by using a pair of complementary mutagenic oligonucleotides in PCR. All of the mutant proteins were highly resistant to the midgut proteases and chymotrypsin. Digestion of the mutant proteins by insect midgut extract and chymotrypsin produced only the active 58-kDa fragment, except that L144H was partially cleaved at residue 144.     

The Mode of Action of the Bacillus thuringiensis Vegetative Insecticidal Protein Vip3A Differs from That of Cry1Ab δ-Endotoxin

The Vip3A protein, secreted by Bacillus spp. during the vegetative stage of growth, represents a new family of insecticidal proteins. In our investigation of the mode of action of Vip3A, the 88-kDa Vip3A full-length toxin (Vip3A-F) was proteolytically activated to an approximately 62-kDa core toxin either by trypsin (Vip3A-T) or lepidopteran gut juice extracts (Vip3A-G). Biotinylated Vip3A-G demonstrated competitive binding to lepidopteran midgut brush border membrane vesicles (BBMV). Furthermore, in ligand blotting experiments with BBMV from the tobacco hornworm, Manduca sexta (Linnaeus), activated Cry1Ab bound to 120-kDa aminopeptidase N (APN)-like and 250-kDa cadherin-like molecules, whereas Vip3A-G bound to 80-kDa and 100-kDa molecules which are distinct from the known Cry1Ab receptors. In addition, separate blotting experiments with Vip3A-G did not show binding to isolated Cry1A receptors, such as M. sexta APN protein, or a cadherin Cry1Ab ecto-binding domain. In voltage clamping assays with dissected midgut from the susceptible insect, M. sexta, Vip3A-G clearly formed pores, whereas Vip3A-F was incapable of pore formation. In the same assay, Vip3A-G was incapable of forming pores with larvae of the nonsusceptible insect, monarch butterfly, Danaus plexippus (Linnaeus). In planar lipid bilayers, both Vip3A-G and Vip3A-T formed stable ion channels in the absence of any receptors, supporting pore formation as an inherent property of Vip3A. Both Cry1Ab and Vip3A channels were voltage independent and highly cation selective; however, they differed considerably in their principal conductance state and cation specificity. The mode of action of Vip3A supports its use as a novel insecticidal agent.     

Rapid Identification of δ-endotoxin from Bacillus thuringiensis subsp. kurstaki HD-1 with Proteomic Analysis

This study was carried out to identify rapidly δ-endotoxin from Bacillus thuringiensis (Bt) subsp. kurstaki HD-1 with proteomic analysis. Protoxin was isolated from sporulated cells and purified by ultracen-trifugation using 40-70% sucrose density gradient. Protoxin was treated with trypsin to analyze digested peptides by liquid chromatographytandem mass spectrometry. The proteomic analysis for detected peptides revealed that this methodology is available for discriminating similar Bt strains by identifying Bt subsp. kurstaki HD-1-specific peptides, suggesting that proteomic analysis can be used for rapid identification of Bt strains.     

Secondary structure of the entomocidal toxin from Bacillus thuringiensis subsp. kurstaki HD-73

The secondary structure of the toxin fromBacillus thuringiensis subsp.kurstaki (Btk) HD-73 was estimated by Raman, infrared, and circular dichroism spectroscopy, and by predictive methods. Circular dichroism and infrared spectroscopy gave an estimate of 33–40% -helix, whereas Raman and predictive methods gave approximately 20%. Raman and circular dichroism spectra, as well as predictive methods, indicated that the toxin contains 32–40% -sheet structure, whereas infrared spectroscopy gave a slightly lower estimate. Thus, all of these approaches are in agreement that the native conformation of Btk HD-73 toxin is highly folded and contains considerable amounts of both -helical and -sheet structures. No significant differences were detected in the secondary structure of the toxin either in solution or as a hydrated pellet.     

Specificity of Activated CryIA Proteins from Bacillus thuringiensis subsp. kurstaki HD-1 for Defoliating Forest Lepidoptera

The insecticidal activity of the CryIA(a), CryIA(b), and CryIA(c) toxins from Bacillus thuringiensis subsp. kurstaki HD-1 was determined in force-feeding experiments with larvae of Choristoneura fumiferana, C. occidentalis, C. pinus, Lymantria dispar, Orgyia leucostigma, Malacosoma disstria, and Actebia fennica. The toxins were obtained from cloned protoxin genes expressed in Escherichia coli. The protoxins were activated with gut juice from Bombyx mori larvae. Biological activity of the individual gene products as well as the native HD-1 toxin was assessed as the dose which prevented 50% of the insects from producing frass within 3 days (frass failure dose [FFD₅₀]). The three toxins were about equally active against M. disstria. In the Choristoneura species, CryIA(a) and CryIA(b) were up to fivefold more toxic than CryIA(c). In the lymantriid species, CryIA(a) and CryIA(b) were up to 100-fold more toxic than CryIA(c). The toxicity of HD-1 was similar to that of the individual CryIA(a) or CryIA(b) toxins in all of these species. None of the CryIA toxins or HD-1 exhibited and toxicity towards A. fennica. Comparison of the observed FFD₅₀ of HD-1 with the FFD₅₀ expected on the basis of its crystal composition suggested a possible synergistic effect of the toxins in the two lymantriid species. Our results further illustrate the diversity of activity spectra of these highly related proteins and provide a data base for studies with forest insects to elucidate the molecular basis of toxin specificity.     

Processing of delta-endotoxin from Bacillus thuringiensis subsp. kurstaki HD-1 and HD-73 by gut juices of various insect larvae.

Midgut juices were prepared from Adoxophyes sp., smaller tea tortrix (STT); Bombyx mori, silkworm (SW); Spodoptera litura, common cutworm (CCW); Plutella xylostella, diamondback moth (DBM); and Musca domestica, housefly (HF) and immobilized onto Sepharose 4B. delta-Endotoxins (ICPs) from Bacillus thuringiensis subsp. kurstaki HD-1 and HD-73 were digested by these immobilized gut juice proteases. All gut juices tested derived relatively proteolytic resistant cores from ICP. The molecular sizes of these cores, about 55 kDa in SDS-PAGE, were resulted. In the case of CCW, however, digestion was very strong and only 1/20 concentration of core protein remained relative to other digests. The N-terminal amino acid sequencing of the core proteins showed that they were truncated at the very end of the N-terminus of protoxin, CryIA, at different sites. Although housefly larvae were completely insensitive to active toxin, the gut juice produced the core, suggesting that the housefly may lack the binding sites for the core-active toxin.     

Formation of Crystalline δ-Endotoxin or Poly-β-Hydroxybutyric Acid Granules by Asporogenous Mutants of Bacillus thuringiensis

Parental strains and asporogenous mutants of Bacillus thuringiensis subspp. kurstaki and aizawai produced high yields of δ-endotoxin on M medium, which contained 330 μg of potassium per ml, but not on ST and ST-a media, each of which contained only 11 μg of potassium per ml. On ST and ST-a media, refractile granules were formed instead. These granules had no insecticidal activity against silkworms and were isolated and identified as poly-β-hydroxybutyric acid. Supplementation of the potassium-deficient ST-a medium with 0.1% KH₂PO₄ (3.7 mM) led to the formation of crystalline δ-endotoxin. The replacement of KH₂PO₄ with equimolar amounts of KCl, KNO₃, and potassium acetate or an equivalent amount of K₂SO₄ had a similar effect, whereas the addition of an equimolar amount of NaH₂PO₄ or NH₄H₂PO₄ did not cause the endotoxin to form. An asporogenous mutant, B. thuringiensis subsp. kurstaki strain 290-1, produced δ-endotoxin on ST-a medium supplemented with 3 mM or more potassium but formed only poly-β-hydroxybutyric acid granules on the media containing ≤1 mM potassium. These results clearly indicate that a certain concentration of potassium is essential for the fermentative production of δ-endotoxin by these isolates of B. thuringiensis. Manganese could not be substituted for potassium. Phosphate ions stimulated poly-β-hydroxybutyric acid formation by strain 290-1. The sporulation of B. thuringiensis and several other Bacillus strains was suppressed on the potassium-deficient ST medium. This suggests that potassium plays an essential role not only in Bacillus cell growth and δ-endotoxin formation but also in sporulation.     

Analysis of Mutations in the Pore-Forming Region Essential for Insecticidal Activity of a Bacillus thuringiensis δ-Endotoxin

The Bacillus thuringiensis insecticidal delta-endotoxins have a three-domain structure, with the seven amphipathic helices which comprise domain I being essential for toxicity. To better define the function of these helices in membrane insertion and toxicity, either site-directed or random mutagenesis of two regions was performed. Thirty-nucleotide segments in the B. thuringiensis cry1Ac1 gene, encoding parts of helix alpha4 and the loop connecting helices alpha4 and alpha5, were randomly mutagenized. This hydrophobic region of the toxin probably inserts into the membrane as a hairpin. Site-directed mutations were also created in specific surface residues of helix alpha3 in order to increase its hydrophobicity. Among 12 random mutations in helix alpha4, 5 resulted in the total loss of toxicity for Manduca sexta and Heliothis virescens, another caused a significant increase in toxicity, and one resulted in decreased toxicity. None of the nontoxic mutants was altered in toxin stability, binding of toxin to a membrane protein, or the ability of the toxin to aggregate in the membrane. Mutations in the loop connecting helices alpha4 and alpha5 did not affect toxicity, nor did mutations in alpha3, which should have enhanced the hydrophobic properties of this helix. In contrast to mutations in helix alpha5, those in helix alpha4 which inactivated the toxin did not affect its capacity to oligomerize in the membrane. Despite the formation of oligomers, there was no ion flow as measured by light scattering. Helix alpha5 is important for oligomerization and perhaps has other functions, whereas helix alpha4 must have a more direct role in establishing the properties of the channel.     

Dissolution and Degradation of Bacillus thuringiensis δ-Endotoxin by Gut Juice Protease of the Silkworm Bombyx mori

The dissolution and degradation of †-endotoxin (crystal) of Bacillus thuringiensis subsp. kurstaki strain HD-1 were investigated. Crystals were dissolved in 0.1 M phosphate-carbonate-NaOH buffer at pH > 12. Swelling of crystals occurred in the buffer between pH 10 and 11, and crystals dissolved in the same buffer supplemented with gut juice protease of the silkworm Bombyx mori. The proteolytic dissolution of crystals occurred after a time lag of several minutes in 0.1 M carbonate-NaOH buffer, pH 10.2. The time lag was not observed when crystals were suspended in the buffer for 30 min before the addition of protease. After the dissolution of the crystals and further degradation of the solubilized protein, the appearance of a toxic protein with a molecular weight of 59,000, designated P-59, was observed. Lower-molecular-weight peptides (less than 40,000) showed no toxicity to the silkworm larvae on feeding. Digestion of the 120,000-dalton subunit of the crystal by gut juice protease also produced P-59. These observations suggest the occurrence of a similar process in vivo, i.e., the swelling of crystals due to the alkalinity of gut juice and the production of P-59, dependent on the hydrolysis of swollen crystals by gut juice protease.     

Disruption of a Cadherin Gene Associated with Resistance to Cry1Ac δ-Endotoxin of Bacillus thuringiensis in Helicoverpa armigera

A laboratory strain (GY) of Helicoverpa armigera (Hübner) was established from surviving larvae collected from transgenic cotton expressing a Bacillus thuringiensis var. kurstaki insecticidal protein (Bt cotton) in Gaoyang County, Hebei Province, People's Republic of China, in 2001. The GYBT strain was derived from the GY strain through 28 generations of selection with activated Cry1Ac delivered by diet surface contamination. When resistance to Cry1Ac in the GYBT strain increased to 564-fold after selection, we detected high levels of cross-resistance to Cry1Aa (103-fold) and Cry1Ab (>46-fold) in the GYBT strain with reference to those in the GY strain. The GYBT strain had a low level of cross-resistance to B. thuringiensis var. kurstaki formulation (Btk) (5-fold) and no cross-resistance to Cry2Aa (1.4-fold). Genetic analysis showed that Cry1Ac resistance in the GYBT strain was controlled by one autosomal and incompletely recessive gene. The cross-resistance pattern and inheritance mode suggest that the Cry1Ac resistance in the GYBT strain of H. armigera belongs to “mode 1,” the most common type of lepidopteran resistance to B. thuringiensis toxins. A cadherin gene was cloned and sequenced from both the GY and GYBT strains. Disruption of the cadherin gene by a premature stop codon was associated with a high level of Cry1Ac resistance in H. armigera. Tight linkage between Cry1Ac resistance and the cadherin locus was observed in a backcross analysis. Together with previous evidence found with Heliothis virescens and Pectinophora gossypiella, our results confirmed that the cadherin gene is a preferred target for developing DNA-based monitoring of B. thuringiensis resistance in field populations of lepidopteran pests.     

Comparative toxicity of the HD-1 and NRD-12 strains of Bacillus thuringiensis subsp. kurstaki to defoliating forest Lepidoptera.

Insecticidal activities of sporulated cultures of the HD-1 and NRD-12 strains of Bacillus thuringiensis subsp. kurstaki were compared against four species of defoliating forest lepidopterans in diet-incorporation assays. There was no difference in LC50 between the two strains to larvae of spruce budworm (Choristoneura fumiferana), gypsy moth (Lymantria dispar), eastern hemlock looper (Lambdina fiscellaria fiscellaria), and whitemarked tussock moth (Orgyia leucostigma), whether expressed as total alkaline soluble protein, activated toxin protein, or International Units as determined by bioassay against Trichoplusia ni. Both strains were consistently more toxic than HD-1-S-1980 when compared on the basis of alkali-soluble protein, but not on the basis of activated toxin or International Units. Hybridization of genomic DNA after restriction with HindIII revealed the presence of all three cryIA toxin genes in each of the isolates used in this study, including HD-1-S-1980, which was previously reported to have lost the cryIA(b) gene.     

Secondary structure of the entomocidal toxin fromBacillus thuringiensis subsp.kurstaki HD-73

The secondary structure of the toxin fromBacillus thuringiensis subsp.kurstaki (Btk) HD-73 was estimated by Raman, infrared, and circular dichroism spectroscopy, and by predictive methods. Circular dichroism and infrared spectroscopy gave an estimate of 33–40% α-helix, whereas Raman and predictive methods gave approximately 20%. Raman and circular dichroism spectra, as well as predictive methods, indicated that the toxin contains 32–40% β-sheet structure, whereas infrared spectroscopy gave a slightly lower estimate. Thus, all of these approaches are in agreement that the native conformation of Btk HD-73 toxin is highly folded and contains considerable amounts of both α-helical and β-sheet structures. No significant differences were detected in the secondary structure of the toxin either in solution or as a hydrated pellet.     

Identification of a δ-Endotoxin Gene Product Specifically Active against Spodoptera littoralis Bdv. among Proteolysed Fractions of the Insecticidal Crystals of Bacillus thuringiensis subsp. aizawai 7.29

At least three different insecticidal crystal protein genes were shown to be expressed in Bacillus thuringiensis subsp. aizawai 7.29, a strain that is potentially active against the cotton leafworm Spodoptera littoralis Bdv. Among crude K-60 fractions (60- to 70-kilodalton [kDa] molecules) that were products of proteolysed crystals containing the active domains of the protoxin molecules, we were able to distinguish several distinct components on the basis of their antigenic relationship and their larvicidal properties. A purified fraction designated SF2 was a 61-kDa component specifically active against Pieris brassicae L. and homologous to the B. thuringiensis subsp. berliner 1715 plasmid-encoded crystal protein. A second fraction designated SF1 was composed of 63- and 65-kDa polypeptides and was specifically active against S. littoralis. The SF1 fraction and particularly the 65-kDa component were not antigenically related to the 61-kDa component. The purified fractions were compared with the products of three different crystal protein genes we previously cloned from total DNA of B. thuringiensis subsp. aizawai, among them a new type of crystal protein gene encoding a protein that is specifically active against S. littoralis and other insects of the Noctuidae family. This approach led us to consider the 65-kDa component as a minimum active part of a δ-endotoxin that is encoded by this new gene. Products of the two other cloned genes can be correlated with the 61- and 63-kDa components, respectively. Thus, in B. thuringiensis subsp. aizawai 7.29, multiple δ-endotoxin genes of different structural types direct the synthesis of several δ-endotoxins with different host specificities which were identified as components of the insecticidal crystals.     

Molecular and Biochemical Characterization of an Endochitinase (ChiA-HD73) from Bacillus thuringiensis subsp. kurstaki HD-73

An endochitinase gene (chiA-HD73) from the insecticidal bacterium Bacillus thuringiensis subsp. kurstaki HD-73 was cloned, sequenced, and expressed in Escherichia coli DH5αF′. The chitinase activity of the encoded protein was studied in assays with different fluorogenic substrates. The chiA-HD73 gene contained an open-reading frame that encoded an endochitinase with a deduced molecular weight and an isoelectric point of, respectively, 74.5 kDa and 5.75. A putative signal peptide with cleavage sites for both Gram-positive and Gram-negative bacteria was identified. Comparison of ChiA-HD73 with other chitinases revealed a modular structure composed of a catalytic domain and a putative chitin-binding domain. ChiA-HD73 hydrolyzed both tetrameric and trimeric fluorogenic substrates, but not a chitobiose analog substrate, suggesting that the activity of ChiA-HD73 is mainly endochitinolytic. In addition, ChiA-HD73 showed high enzymatic activity within a broad pH range (pH 4–10), with a peak activity at pH 6.5. The optimal temperature for enzymatic activity was observed at 55°C. Its activity in a broad range of temperatures and pH suggests ChiA-HD73 could have biotechnological applications in insect control, particularly in synergizing the insecticidal crystal protein toxins of B. thuringiensis.     

Comparative analysis of the individual protoxin components in P1 crystals of Bacillus thuringiensis subsp. kurstaki isolates NRD-12 and HD-1.

Two commercially important strains (NRD-12 and HD-1) of the entomopathogenic bacterium Bacillus thuringiensis subsp. kurstaki each contain three genes of partially identical sequence coding for three classes of 130-135 kDa protoxins (termed the 4.5, 5.3 and 6.6 protoxins) that display toxicity towards various lepidopteran larvae. These gene products combine to form the intracellular bipyramidal P1 crystal. Each of the genes from both strains was cloned and expressed in Escherichia coli. Analysis of the cloned genes at the restriction-endonuclease level revealed no detectable differences among genes within a particular gene class. The composition of the P1 crystal from both strains was quantitatively analysed by CNBr cleavage of the purified P1 crystal, with the purified recombinant gene products as reference proteins. Independent verification of the presence of high 6.6-protoxin gene product in the P1 crystal was provided by a rapid in vitro lawn cell toxicity assay directed against a Choristoneura fumiferana (CF-1) insect cell line. The results indicate that, although all three gene products are represented within the P1 crystal of either NRD-12 or HD-1, only the contents of the 4.5 and 5.3 protoxins vary between the two crystals, whereas the 6.6 protoxin contents are similar and represent approximately one-third of the P1 crystal in either strain.     

Toxicity of chitinase-producing Bacillus thuringiensis ssp. kurstaki HD-1 (G) toward Plutella xylostella

One-hundred fifty isolates of Bacillus thuringiensis were tested for their ability to produce chitinase using colloidal chitin agar as the primary plating medium. Of 14 strains that produced chitinase, B. thuringiensis ssp. kurstaki HD-1(G) was identified as the highest chitinase producer and selected for further study. This bacterium produced the highest amount of chitinase (19.3 mU/ml) when it was cultivated in nutrient broth supplemented with 0.3% colloidal chitin on a rotary shaker (200 rpm) at 30 degrees C for 2 days. The toxicities of B. thuringiensis ssp. kurstaki HD-1(G) and B. thuringiensis ssp. kurstaki wa-p-2, a chitinase nonproducer, were assayed toward Plutella xylostella (diamondback moth) larvae, resulting in LC(50)'s of 4.93 x 10(4) and 1.32 x 10(5) spores/ml, respectively. If the culture broth from B. thuringiensis ssp. kurstaki HD-1(G) was used as the suspending liquid instead of phosphate buffer, their LC(50)'s were reduced to 6.23 x 10(3) and 7.60 x 10(4) spores/ml, respectively. The histopathological changes of the midgut epithelial cells of diamondback moth larvae were compared after feeding on B. thuringiensis ssp. kurstaki HD-1(G) with and without the presence of supernatant containing chitinase under light microscopy and transmission electron microscopy. The midgut epithelial cells of larvae fed for 30 min in the presence of chitinase, with or without spores and endotoxin crystals, appeared more elongated and swollen than those of the control larvae. A number of different cellular changes such as extensive cellular disintegration and appearance of numerous vacuoles were observed from the larvae fed on B. thuringiensis ssp. kurstaki HD-1(G) supplemented with supernatant containing chitinase. Thus increased toxicity and changes in epithelial cells were correlated with the presence of chitinase but this was not distinguished from the possible presence of vegetative-stage insecticidal proteins.