Fluorometric Assay of Potential Change of Bombyx mori Midgut Brush Border Membrane Induced by δ-Endotoxin from Bacillus thuringiensis

Bacillus thuringiensis δ-endotoxin, CryIA(a), increased ion permeability of brush border membrane vesicles isolated from midgut epithelia of Bombyx mori larvae. This ion permeability change was measured with a potential-sensitive fluorescent dye, 3,3′-dipropylthiadicarbocyanine iodide. This effect was observed at concentrations of the toxin that correspond to normal effective doses in vivo. CryIA(b) and heat-treat CryIA(a) did not show this effect. CryIA(a) did not show this effect on rat renal brush border membranes. The effects depended on the toxin dose in saturable manner. These suggest that this assay system reflects the mode of action of δ-endotoxin in vivo. The toxin increased various ion permeabilities, suggesting that δ-endotoxin forms non-selective cation pores on brush border membranes.     

Toxicity and Receptor Binding Properties of Bacillus thuringiensisδ-Endotoxins to the Midgut Brush Border Membrane Vesicles of the Rice Leaf Folders, Cnaphalocrocis medinalis and Marasmia patnalis

Pesticidal activity and receptor-binding properties of Bacillus thuringiensis toxins to rice leaf folders, Cnaphalocrocis medinalis and Marasmia patnalis, were investigated. Saturation and competition binding experiments were done with iodine (¹²⁵1)-labeled Bt proteins and brush border membrane vesicles prepared from the midgut of C. medinalis and M. patnalis. The results show saturable, specific, and high-affinity binding of all toxins except Cry2A toxin. Cry1Aa and Cry2A toxins were bound with low affinity but with high binding site concentration. Heterologous competition experiments showed that Cry1Aa, Cry1Ab, and Cry1Ac recognized or shared the same binding site that is different from the binding site for Cry2A toxin. Iodine (¹²⁵I)-labeled Cry1Ac and Cry1Ab toxins were used in ligand blot experiments to detect specific binding proteins in brush border membrane vesicles of C. medinalis and M. patnalis. Cry1Ab toxin protein binds to 205-kDa and 200-kDa proteins respectively in case of C. medinalis and M. patnalis. The apparent molecular mass of the protein bound to labeled Cry1Ac toxins was identified as a 120-kDa protein in both C. medinalis and M. patnalis. Received: 10 April 2000 / Accepted: 23 May 2000     

Improvement of Bacillus thuringiensis δ-endotoxins Synthesis Yields Through Acquisition of Erythromycin Resistance

The acquisition of the erythromycin resistance by Bacillus thuringiensis kurstaki improved yields of δ-endotoxins in sporulating cells ranging from 134 to 215%. Resistance to erythromycin decreased the final spore count by at least 50%. Consequently, erythromycin resistance is an efficient tool for the improvement of bioinsecticides yields with a high ratio of δ-endotoxins to spores. Revisions requested 31 October 2005; Revisions received 28 November 2005     

Binding of Cry δ-endotoxins to brush border membrane vesicles of Helicoverpa armigera and Helicoverpa punctigera （Lepidoptera： Noctuidae）

The relatively low susceptibility ofHelicoverpa armigera to CrylAc, its history of resistance to chemical insecticides and the seasonal decline in expression of CrylAc in transgenic cotton necessitated the development of cotton expressing two insecticidal proteins to provide sustainable control of this multinational pest. To manage the resistance issue, it was essential that the second insecticidal protein have a significantly different mode of action to CrylAc. A common feature of resistance to CrylA proteins in several species as well as H. armigera has been a change in the binding site. A study of binding sites for some Cry proteins in the brush border membrane vesicles （BBMV） ofH. armigera and Helicoverpa punctigera was undertaken. The binding affinity for CrylAc was higher than for CrylAb, matching their relative toxicities, and CrylAc and CrylAb were found to share at least one binding site in both I-1. armigera and I-1. punctigera. However Cry2Aa did not compete with CrylAc for binding and so could be used in transgenic cotton in combination with CrylAc to control H. armigera and manage resistance. Variation in the susceptibilities of three different H. armigera strains to CrylAc correlated with the parameter Bmax/Kcom.     

Colloid-osmotic lysis is a general feature of the mechanism of action of Bacillus thuringiensis δ-endotoxins with different insect specificity

The mechanism of action of Bacillus thuringiensis insecticidal δ-endotoxins has long been the subject of controversy. As our working hypothesis we propose a two-step model in which, after binding a specific plasma membrane receptor, the action of all the δ-endotoxins studied here is to generate small pores in the plasma membrane, either directly by inserting into the membrane, or indirectly by perturbing resident plasma membrane molecules. The creation of these pores will lead to colloid-osmotic lysis, i.e., an equilibration of ions through the pore resulting in a net inflow of ions, an accompanying influx of water, cell swelling and eventual lysis. Our observations that cell swelling precedes lysis, that small molecules leak out of the cell before large ones, that osmotic protectants inhibit or delay cytolysis, and that the toxin-induced pore of 0.5–1.0 nm radius will allow equilibration of ions but not leakage of cytoplasmic macromolecules, are in full agreement with the predictions of this hypothesis. To explain the specificity of the δ-endotoxin-induced lytic pore formation, we propose that prior interaction between the toxin and cell-specific plasma membrane recpetors is necessary before these toxins can insert into, or interact with, the membrane.     

Two Different Bacillus thuringiensis Delta-Endotoxin Receptors in the Midgut Brush Border Membrane of the European Corn Borer, Ostrinia nubilalis (Hübner) (Lepidoptera: Pyralidae)

Binding of three Bacillus thuringiensis insecticidal crystal proteins (ICPs) to the midgut epithelium of Ostrinia nubilalis larvae was characterized by performing binding experiments with both isolated brush border membrane vesicles and gut tissue sections. Our results demonstrate that two independent ICP receptors are present in the brush border of O. nubilalis gut epithelium. From competition binding experiments performed with I-labeled and native ICPs it was concluded that CryIA(b) and CryIA(c) are recognized by the same receptor. An 11-fold-higher binding affinity of CryIA(b) for this receptor correlated with a 10-fold-higher toxicity of this ICP compared with CryIA(c). The CryIB toxin did not compete for the binding site of CryIA(b) and CryIA(c). Immunological detection of ingested B. thuringiensis ICPs on gut sections of O. nubilalis larvae revealed binding only along the epithelial brush border membrane. CryID and CryIE, two ICPs that are not toxic to O. nubilalis, were not bound to the apical microvilli of gut epithelial cells. In vitro binding experiments performed with native and biotinylated ICPs on tissue sections confirmed the correlation between ICP binding and toxicity. Moreover, by performing heterologous competition experiments with biotinylated and native ICPs, it was confirmed that the CryIB receptor is different from the receptor for CryIA(b) and CryIA(c). Retention of activated crystal proteins by the peritrophic membrane was not correlated with toxicity. Furthermore, it was demonstrated that CryIA(b), CryIA(c), and CryIB toxins interact in vitro with the epithelial microvilli of Malpighian tubules. In addition, CryIA(c) toxin also adheres to the basement membrane of the midgut epithelium.     

An Effect of Ca2+ on the Intrinsic Cl−-conductance of Rat Kidney Cortex Brush Border Membrane Vesicles

Brush-border membrane vesicles (BBMV) were prepared from superficial rat renal cortex by a divalent²⁺-precipitation technique using either CaCl₂ or MgCl₂. The dependence of the initial [¹⁴C]-d-glucose (or [³H]-l-proline) uptake rate and the extent of the overshoot of d-glucose or l-proline uphill accumulation from solutions containing 100 mm Na⁺ salt, was found to be dependent upon the precipitating divalent cation. With Mg²⁺ precipitation the initial uptake and overshoot accumulation of either d-glucose or l-proline were enhanced compared to BBMV prepared by Ca²⁺ precipitation. When the anion composition of the media was varied (uptake in Cl⁻ media in comparison to gluconate⁻-containing media) it was found that the Cl⁻-dependent component of the initial uptake was markedly depressed with Ca²⁺-prepared BBMV (104.99 ± 33.31 vs. 13.83 ± 1.44 pmoles/sec/mg protein for Mg²⁺ and Ca²⁺ prepared vesicles respectively). When Ca²⁺ was loaded into Mg²⁺ prepared BBMV using a freeze-thaw technique, it was found that the magnitude and Cl⁻ enhancement of d-glucose transport was reduced in a dose-dependent manner. Neomycin, an inhibitor of phospholipase C, had no effect on the reduction of d-glucose uptake by Ca²⁺ in Mg²⁺ prepared vesicles. In contrast, phosphatase inhibitors such as vanadate and fluoride were able to partially reverse the Ca²⁺ inhibition of d-glucose uptake and restore the enhancement due to Cl⁻ media. In addition, inhibitors of protein phosphatase 2B, deltamethrin (50 nm) and trifluoperazine (10 μm), caused partial reversal of Ca²⁺-dependent inhibition of d-glucose uptake. Direct measurement of changes in the bi-ionic (Cl⁻ vs. gluconate⁻) transmembrane electrical potential differences using the cyanine dye, 3,3′-dipropylthiodicarbocyanine iodide DiSC₃-(5) confirmed that Cl⁻ conductance was reduced in Ca²⁺-prepared vesicles. We conclude that a Cl⁻ conductance coexists with Na⁺ cotransport in rat renal BBMV and this may be subject to negative regulation by Ca²⁺ via stimulation of protein phosphatase (PP2B). Received: 14 December 1994/Revised: 27 November 1995     

Carboxy-terminal extension effects on crystal formation and insecticidal properties of Colorado potato beetle-active Bacillus thuringiensis δ-endotoxins

Many Bacillus thuringiensis crystal proteins, particularly those active against lepidopteran insects, have carboxy-terminal extensions that mediate bipyramidal crystal formation. These crystals are only soluble at high (>10.0) pH in reducing conditions such as generally found in the lepidopteran midgut. Most of the Colorado potato beetle (CPB)-active toxins lack such an extension, yet some toxins with a carboxy-terminal extension have cryptic activity against this insect, revealed only after in vitro solubilization. Crystal formation, morphology, protein content, and activity against CPB were compared for two sets of proteins, the Cry 1-hybrid SN19 and Cry3Aa, both with and without a carboxy-terminal extension. Cry3Aa, with or without extension, formed flat square or rectagular crystals. SN19 (with extension) and its derivative without extension formed irregular inclusion bodies. All Cry3Aa and SN19 crystals and inclusion bodies were almost equally active before and after in vitro presolubilization and could be solubilized in diluted CPB midgut extract. In contrast, bipyramidal crystals of Cry1Ba were insoluble under these conditions. Our results suggest that bipyramidal crystal formation typical for proteins with a carboxy-terminal extension may preclude activity against CPB, but that interfering with this crystal formation can increase the activity.     

Influence of some plant phenolics on the activity of δ-endotoxin of Bacillus thuringiensis var. galleriae on Heliothis armigera

Bioassays with Bacillus thuringiensis var. galleriae Berliner -endotoxin and plant phenolics on Heliothis armigera Hübner enhanced the activity of B.t. var. galleriae endotoxin. The presence of plant phenolics with B.t var. galleria endotoxin not only reduced feeding potential and weight gain by the larvae, but also enhanced the LC50 value of the toxin. Our study demonstrates the effect of phytochemicals from resistant crop plants on the biocidal activity of B. thuringiensis strains in laboratory conditions.     

The main features of Bacillus thuringiensis δ-endotoxin molecular structure

The crystal-forming proteins (-endotoxins) produced by various serotypes of Bacillus thuringiensis and toxic for Lepidoptera reveal the same pattern of molecular organisation. These proteins (M _r of ca. 145,000–130,000) contain an N-terminal domain (M _r of 85,000–65,000) resistant to proteolysis whereas their C-terminal moieties (M _r of 65,000) undergo an extensive degradation by trypsin that leads to stepwise cleavage off the fragments with M _r of 15,000–35,000.The N-terminal domain from serotype V -endotoxin is active when introduced into the hemocoel of Galleria mellonella larvae. Hence, it correponds to the true toxin normally formed by larva proteases action on the crystalforming protein (protoxin). Some differences were found in the properties of the N-terminal domains isolated from the crystal-forming proteins of III, V and IX serotypes, e.g., in their solubility, digestion by subtilisin, molecular weights and the distribution of methionine residues along the polypeptide chains.Abbreviations SDS sodium dodecyl sulphate - PAGE polyacryl amide gel electrophoresis - CFP crystal-forming protein - DNS 5-dimethylamino-1-naphthalene-sulphonyl     

Activity spectra of Bacillus Thuringiensis δ-endotoxins against eight insect cell lines

Summary Eight continuous insect cell lines were tested for susceptibility to the δ-endotoxins of several lepidopteran-active strains and cloned-gene products of Bacillus thuringiensis. The assays were performed on cells suspended in agarose gel, which allowed the toxins activated at pH 10.5 to be applied directly in a high-pH buffer without causing solvent toxicity to the cells. The responses of the cell lines to the various toxins produced activity spectra that were used to identify functionally similar and dissimilar toxin proteins. IPRI-CF-1 and FPMI-MS-5, derived from neonate larvae of Choristoneura fumiferana and Manduca sexta, respectively, exhibited the greatest sensitivity to the toxins tested, whereas B. thuringiensis subsp. entomocidus had the broadest in vitro host range. Analysis of activity spectra led to the identification of the particular Cry protein that was responsible for the broad toxicity of this subspecies and demonstrated a distinct difference in toxin composition between two strains of subsp. sotto. The identical spectra observed for subsp. kurstaki HD-1 and NRD-12 is consistent with insect bioassay data obtained previously by other workers and supports the conclusion that there is virtually no difference in activity between these two strains. The in vitro assay system, referred to as the “lawn assay” and used to test B. thuringiensis activated toxins against insect cell lines, is particularly useful in mode-of-action studies and as a rapid, preliminary test for the presence of specific cytolytic proteins, rather than as a method for screening toxins of wild-type strains for insecticidal activity. The response of cells in vitro to B. thuringiensis toxins is often very different from that of the insect from which the cells were derived.     

Bacillus thuringiensis CrylAa δ-Endotoxin Affects the K+/Amino Acid Symport in Bombyx mori Larval Midgut

The mechanical properties of brush border membrane vesicles, BBMV, from rabbit kidney proximal tubule cells, were studied by measuring the initial and final equilibrium volumes of vesicles subjected to different osmotic shocks, using cellobiose as the impermeant solute in the preparation buffer. An elevated intracellular hydrostatic pressure was inferred from osmotic balance requirements in dilute solutions. For vesicles prepared in 18 and 85 mosm solutions, these pressures are close to 17 mosm (290 mm Hg). The corresponding membrane surface tension is 6.0 × 10⁻⁵ N cm⁻¹ while the membrane surface area is expanded by at least 2.2%. When these vesicles are exposed to very dilute solutions the internal hydrostatic pressure rises to an estimated 84 mosm (1444 mm Hg) just prior to lysis. The corresponding maximal surface tension (pre-lysis) is 18.7 × 10⁻⁵ N cm⁻¹, and the maximal expansion of membrane area is 6.8%. The calculated area compressibility elastic modulus was 2.8 × 10⁻³ N cm⁻¹. Received: 8 August 1996/Revised: 4 March 1997     

Susceptibility of major lepidopterans to δ-endotoxins and a formulation of Bacillus thuringiensis (B.t.) on vegetable brassicas in Taiwan

Baseline susceptibility of Plutella xylostella, Crocidolomia binotalis and Hellula undalis to Bacillus thuringiensis (B.t.) δ-endotoxins (Cry1Aa, Cry1Ab, Cry1Ac, Cry1Ca) and a formulation (Xentari®) was assessed using a leaf-dip bioassay method. The toxins Cry1Ac, Cry1Aa and Cry1Ca were equally toxic to P. xylostella. Crocidolomia binotalis was highly susceptible to all Cry1A toxins and it was least sensitive to Cry1Ca. Conversely, H. undalis was highly sensitive to Cry1Ca, but less susceptible to Cry1A toxins. Hence, the susceptibility of H. undalis to Xentari®, a formulation containing Cry1C as a major toxin, was significantly higher than C. binotalis or P. xylostella.     

Importance of spores,crystals, and δ-endotoxins in the pathogenicity of different varieties of Bacillus thuringiensis in Galleria mellonella and Pieris brassicae

Physical methods were used to produce spores containing impurities of 0.02–0.05% crystals and crystals containing impurities of 0.001–0.01% spores from cultures of Bacillus thuringiensis. In Galleria mellonella larvae, these preparations from varieties galleriae, aizawai, and wuhanensis were only moderately active compared to 1:1 mixtures of spores and crystals. Spores of an acrystalliferous aizawai mutant were inactive and did not contain a polypeptide of the same size as the potent M_r 138000 δ-endotoxin present in spores and crystals of all three wild-type strains. Thus, this polypeptide probably contributed to the moderate activity of wild-type spores. Spore impurities in the crystal preparation were killed by γ irradiation without harming the crystals. The crystals without live spores were virtually inactive (LC₅₀s, ca. 10¹⁰ crystals/g insect food). Addition of 10³ spores to 10⁸ crystals/g food (0.001% spores) increased the mortality of larvae from 0 to 36%, and addition of 10⁴ spores (0.01% spores) killed 64% larvae. Thus, the addition of low levels of spores increased the potency of crystals in G. mellonella from virtually zero to moderate levels, suggesting that the live spore impurities in the crystal preparations were responsible for the observed moderate potency of crystals before γ irradiation, a view supported by a reduction of potency of crystal preparations following admixture of streptomycin to the insect food. In contrast to the results with G. mellonella, crystals were ca. 30 times as active as spores in Pieris brassicae larvae. Many authors have found crystals purified by physical methods to be highly active in a range of lepidopterous hosts. The present work indicates that the role of the spore impurities in these species may need further investigation. Absence of live spores of B. thuringiensis may impair the control of some insect species feeding on spore-free products and on microorganisms or plants into which endotoxins have been introduced by genetic manipulation.     

Investigation of the membrane lesion induced in vitro by two mosquitocidal δ-endotoxins ofBacillus thuringiensis

Bacillus thuringiensis produces insecticidal crystalline -endotoxins during sporulation. Leakage of radiolabeled markers from insect tissue culture cells and erythrocytes supports a mechanism of colloid osmotic lysis for the 23K and 27K polypeptides from the -endotoxins ofBacillus thuringiensis var.darmstadiensis 73-E10-2 and var.israelensis, respectively. Both toxins produce a plasma membrane pore 0.6–1.0 nm in radius.     

Variation in susceptibility of Helicoverpa armigera (Hübner) and Helicoverpa punctigera (Wallengren) (Lepidoptera: Noctuidae) in Australia to two Bacillus thuringiensis toxins

Intra-specific variation in susceptibility of Helicoverpa armigera (Hübner) and Helicoverpa punctigera (Wallengren) in Australia to the Cry1Ac and Cry2Ab delta-endotoxins from Bacillus thuringiensis (Berliner) (Bt) was determined to establish a baseline for monitoring changes that might occur with the use of Bt cotton. Strains of H. armigera and H. punctigera were established from populations collected primarily from commercial farms throughout the Australian cotton belts. Strains were evaluated for susceptibility using two bioassay methods (surface treatment and diet incorporation) by measuring the dose response for mortality (LC50) and growth inhibition (IC50). The variation in LC50 among H. armigera (n=17 strains) and H. punctigera (n=12 strains) in response to Cry1Ac was 4.6- and 3.2-fold, respectively. The variation in LC50 among H. armigera (n=19 strains) and H. punctigera (n=12 strains) to Cry2Ab was 6.6- and 3.5-fold, respectively. The range of Cry1Ac induced growth inhibition from the 3rd to 4th instar in H. armigera (n=15 strains) was 3.6-fold and in H. punctigera (n=13 strains) was 2.6-fold, while the range of Cry2Ab induced growth inhibition from neonate to 3rd instar in H. armigera (n=13 strains) was 4.3-fold and in H. punctigera (n=12 strains) was 6.1-fold. Variation in susceptibility was also evaluated for two age classes (neonates and 3rd instars) in laboratory strains of H. armigera and H. punctigera. Neonates of H. punctigera had the same or higher sensitivity to Bt than 3rd instars. Neonates of H. armigera were more sensitive to Cry2Ab than 3rd instars, while being less sensitive to Cry1Ac than 3rd instars. Differences in the two methods of bioassay used affected relative sensitivity of species to Bt toxins, highlighting the need to standardize bioassay protocols.     

Effect of sublethal concentration of Bacillus thuringiensis var. kurstaki on food and developmental needs of the american bollworm, Helicoverpa armigera (Hübner)

Effect of sublethal concentration of B. thuringiensis on the first, third, fourth and fifth instar larvae of the American bollworm, H. armigera was investigated to study their response to food consumption, digestion, utilization, and their development till adult formation. The young larvae surviving B. thuringiensis treatment in their first instar and third instar delayed larval period by two to three days, but did not consume more food as compared to control. However, they showed higher digestibility of food as compared to control, which was compensated by their reduced ability to utilize the digested food for body substance. Contrary to the effect on first and third instar larvae, the fifth instar larvae surviving B. thuringiensis treatment in its fourth instar consumed less food, showed less absorption efficiency in digesting food, but compensated by increase in the utilization of ingested and digested food into body substance. Insects surviving B. thuringiensis HD-1 sublethal toxicity adapted to normal larval growth when fed on untreated food, depending upon insect growth prior to treatment. The moths emerging from B.thuringiensis treated larvae had sex ratio favouring females, and adult pairs laid less fertile eggs than those from the untreated ones. The response of B. thuringiensis treated larvae to their food and developmental needs is discussed.     

Binding of Insecticidal Crystal Proteins of Bacillus thuringiensis to the Midgut Brush Border of the Cabbage Looper, Trichoplusia ni (Hübner) (Lepidoptera: Noctuidae), and Selection for Resistance to One of the Crystal Proteins

The susceptibility of Trichoplusia ni larvae to several Bacillus thuringiensis insecticidal crystal proteins (ICPs) was tested. Neonatal larvae proved to be susceptible to solubilized trypsin-treated CryIA(a), CryIA(b), and CryIA(c) (50% lethal concentrations [LC(50)s], 570, 480, and 320 ng/cm, respectively) but showed little susceptibility to CryIB and CryID (LC(50)s, 5,640 and 2,530 ng/cm, respectively). The toxicity of ICPs was correlated to binding to the epithelial brush border of the midgut, as revealed by immunocytochemical staining with monoclonal antibodies. In vitro binding experiments with iodinated ICPs and brush border membrane vesicles indicated that CryIA(b) and CryIA(c) share the same high-affinity binding site, whereas CryIA(a) binds to a different one. The affinities of CryIA(b) and CryIA(c) for the binding site were similar (K(d) = 3.6 and 4.7 nM, respectively), and the mean binding-site concentration was 0.71 pmol/mg of vesicle protein. Selection of a population with increasing concentrations of CryIA(b) produced 31-fold resistance in seven generations. The realized heritability (h) was 0.19. The increase of homozygosity (for resistance factors) as selection proceeded was reflected in the increase in the slopes of the dose-mortality curves. Resistance was specific for CryIA(b) and did not extend to CryIA(a) or even to CryIA(c). This result was not predicted by the binding-site model, in which CryIA(b) and CryIA(c) bind to the same high-affinity binding site. This result may suggest a more complicated relationship between in vitro binding of ICPs to specific sites in the epithelial membrane of the midgut and the in vivo toxic effect.     

Development and mechanisms of resistance to Bacillus thuringiensis endotoxin Cry1Ac in the American bollworm, Helicoverpa armigera (Hübner)

The American bollworm, H. armigera, evolved 31-fold resistance to selection pressure of B. thuringiensis endotoxin Cry1Ac within six generations. The Cry1Ac selected larvae of H. armigera showed cross-resistance to Cry1Aa and Cry1Ab both in terms of mortality and growth reduction. Studies on mechanisms of resistance to Cry1Ac showed that proteases of resistant insects degraded Cry1Ac faster than those of susceptible insects, which led to the relative unavailability of toxin of about 58 kDa for binding and perforation of midgut epithelial membrane of the target insect. Besides, resistant and susceptible populations of H. armigera differed in the binding of their receptors with Cry1Ac toxin. These results suggest the possibility of both mechanisms existing in imparting resistance. These findings mandate the necessity of B. thuringiensis resistance management for usage of B. thuringiensis either as a conventional insecticide or through transgenic crops.     

Potentiation of insecticidal activity of Bacillus thuringiensis subsp. kurstaki HD-1 by proteinase inhibitors in the American bollworm, Helicoverpa armigera (Hübner)

The effect of crude proteinase inhibitor extracts from seeds of different crop plants (black gram, chickpea, chickling vetch, finger millet, French bean, green gram, horse gram, lentil, pea and soybean) on the insecticidal activity of B. thuringiensis var. kurstaki HD-1 was investigated against neonate larvae of H. armigera by diet incorporation method. The larval mortality due to crude proteinase inhibitors alone (5% seed weight equivalent) ranged from 4.1 to 19.1%; the maximum mortality with finger millet and the minimum with pea var. DDR-23. A mixture of B. thuringiensis var. kurstaki HD-1 (10 ppm) and proteinase inhibitor (5% seed weight equivalent) was synergistic in larval mortality with respect to proteinase inhibitors of pea var. DMR-16, chickling vetch var. RLK-1098 and B101-212, lentil var. ILL-8095 and L-4076, soybean var. PK-1042, PK-416 and Pusa-22, chickpea var. Pusa-413, French bean (Chitra) and black gram; and antagonistic with respect to those of finger millet, horse gram and kidney bean. The larval growth reduction with crude proteinase inhibitors alone ranged from 17.9 to 53.1%; the maximum growth reduction with soybean var. PK-1042 and minimum with lentil var. L-4076. A mixture of B. thuringiensis var. kurstaki and proteinase inhibitor was synergistic in growth reduction with respect to proteinase inhibitors of lentil var. ILL-8095, and L-4626 and antagonistic with respect to that of finger millet. The midgut proteinase inhibition with crude seed extracts (3.3% seed weight equivalent) ranged from 9.3 to 60.9% and was negatively correlated with larval mortality. These results showed that interactive effect of B. thuringiensis var. kurstaki HD-1 and proteinase inhibitors in the larvae of H. armigera depended upon the quality and quantity of proteinase inhibitors, which vary widely in different plants.