Toxicity and mode of action of insecticidal Cry1A proteins from Bacillus thuringiensis in an insect cell line, CF-1

Cry4Aa and Cry4Ba Mosquito-Active Toxins Utilize Different Domains in Binding to a Particular Culex ALP Isoform: A Functional Toxin Receptor Implicating Differential Actions on Target Larvae

The three-domain Cry4Aa toxin produced from Bacillus thuringiensis subsp. israelensis was previously shown to be much more toxic to Culex mosquito larvae than its closely related toxin—Cry4Ba. The interaction of these two individual toxins with target receptors on susceptible larval midgut cells is likely to be the critical determinant in their differential toxicity. Here, two full-length membrane-bound alkaline phosphatase (mALP) isoforms from Culex quinquefasciatus larvae, Cq-mALP1263and Cq-mALP1264, predicted to be GPI-linked was cloned and functionally expressed in Spodoptera frugiperda (Sf9) cells as 57- and 61-kDa membrane-bound proteins, respectively. Bioinformatics analysis disclosed that both Cq-mALP isoforms share significant sequence similarity to Aedes aegypti-mALP—a Cry4Ba toxin receptor. In cytotoxicity assays, Sf9 cells expressing Cq-mALP1264, but not Cq-mALP1263, showed remarkably greater susceptibility to Cry4Aa than Cry4Ba, while immunolocalization studies revealed that both toxins were capable of binding to each Cq-mALP expressed on the cell membrane surface. Molecular docking of the Cq-mALP1264-modeled structure with individual Cry4 toxins revealed that Cry4Aa could bind to Cq-mALP1264 primarily through particular residues on three surface-exposed loops in the receptor-binding domain—DII, including Thr⁵¹², Tyr⁵¹³ and Lys⁵¹⁴ in the β10-β11loop. Dissimilarly, Cry4Ba appeared to utilize only certain residues in its C-terminal domain—DIII to interact with such a Culex counterpart receptor. Ala-substitutions of selected β10-β11loop residues (T512A, Y513A and K514A) revealed that only the K514A mutant displayed a drastic decrease in biotoxicity against C. quinquefasciatus larvae. Further substitution of Lys⁵¹⁴ with Asp (K514D) revealed a further decrease in larval toxicity. Furthermore, in silico calculation of the binding affinity change (ΔΔG_bind) in Cry4Aa-Cq-mALP1264 interactions upon these single-substitutions revealed that the K514D mutation displayed the largest ΔΔG_bind value as compared to three other mutations, signifying an adverse impact of a negative charge at this critical receptor-binding position. Altogether, our present study has disclosed that these two related-Cry4 mosquito-active toxins conceivably exploited different domains in functional binding to the same Culex membrane-bound ALP isoform—Cq-mALP1264 for mediating differential toxicity against Culex target larvae.

Nitrogen-Doped Carbon Dots as a Fluorescent Probe for the Highly Sensitive Detection of Bilirubin and Cell Imaging

Nitrogen-doped carbon dots (NCDs) with bright blue fluorescence were constructed by a hydrothermal method using sucrose and L-proline as raw materials. The NCDs were characterized by transmitted electron microscopy, X-ray diffractometry, Fourier transform infrared spectrometry, X-ray photoelectron spectroscopy, and ultraviolet-visible absorption (UV-vis) and fluorescence spectroscopy to investigate the morphology, elemental composition, and optical properties. The NCDs had good water solubility, high dispersibility with an average diameter of only 1.7 nm, and satisfactory optical properties with a fluorescence quantum yield of 23.4%. The NCDs were employed for the detection of bilirubin. A good linear response of the NCDs in the range 0.35-9.78 μM was obtained for bilirubin with a detection limit of 33 nM. The NCDs were also applied to the analysis of real samples, serum and urine, with a recovery of 95.34%-104.66%. The low cytotoxicity and good biocompatibility of the NCDs were indicated by an MTT assay and cell imaging of HeLa cells. Compared with other detection systems, using NCDs for bilirubin detection was a facile and efficient method with good selectivity and sensitivity.

Cry1Ac Protoxin and Its Activated Toxin from Bacillus thuringiensis Act Differentially during the Pathogenic Process

Although the new dual model of the Bacillus thuringiensis insecticidal mechamism indicated that both Cry1A protoxin and activated toxin have the potency to kill insects, the difference in the toxic pathways elicited by the protoxin and activated toxin was less understood at the molecular level. Through utilizing the CF-203 cell line derived from the midgut of Choristoneura fumiferana, we found that there existed obvious differences in the binding sites and endocytosis pathways for the two forms of Cry1Ac. In addition, it was revealed that Cry1Ac protoxin existed predominantly in the midgut of Plutella xylostella at the early stage after ingesting Cry1Ac crystals, which brought about obvious damage to the midgut epithelium and exhibited different binding sites on the brush border membrane vesicle compared to the toxin. These findings supported the dual mode of action of B. thuringiensis Cry1A proteins and improved our understanding of the molecular features that contribute to the protoxin toxicity.

Study of the Bacillus thuringiensis Cry1Ia Protein Oligomerization Promoted by Midgut Brush Border Membrane Vesicles of Lepidopteran and Coleopteran Insects, or Cultured Insect Cells

Bacillus thuringiensis (Bt) produces insecticidal proteins that are either secreted during the vegetative growth phase or accumulated in the crystal inclusions (Cry proteins) in the stationary phase. Cry1I proteins share the three domain (3D) structure typical of crystal proteins but are secreted to the media early in the stationary growth phase. In the generally accepted mode of action of 3D Cry proteins (sequential binding model), the formation of an oligomer (tetramer) has been described as a major step, necessary for pore formation and subsequent toxicity. To know if this could be extended to Cry1I proteins, the formation of Cry1Ia oligomers was studied by Western blot, after the incubation of trypsin activated Cry1Ia with insect brush border membrane vesicles (BBMV) or insect cultured cells, using Cry1Ab as control. Our results showed that Cry1Ia oligomers were observed only after incubation with susceptible coleopteran BBMV, but not following incubation with susceptible lepidopteran BBMV or non-susceptible Sf21 insect cells, while Cry1Ab oligomers were persistently detected after incubation with all insect tissues tested, regardless of its host susceptibility. The data suggested oligomerization may not necessarily be a requirement for the toxicity of Cry1I proteins.

Cytotoxicity and binding profiles of activated Cry1Ac and Cry2Ab to three insect cell lines

While Cry1Ac has been known to bind with larval midgut proteins cadherin, APN (amino peptidase N), ALP (alkaline phosphatase) and ABCC2 (adenosine triphosphate-binding cassette transporter subfamily C2), little is known about the receptors of Cry2Ab. To provide a clue to the receptors of Cry2Ab, we tested the baseline cytotoxicity of activated Cry1Ac and Cry2Ab against the midgut and fat body cell lines of Helicoverpa zea and the ovary cell line of Spodoptera frugiperda (SF9). As expected, the descending order of cytotoxicity of Cry1Ac against the three cell lines in terms of 50% lethal concetration (LC₅₀ ) was midgut (31.0 μg/mL) > fat body (59.0 μg/mL) and SF9 cell (99.6 μg/mL). By contrast, the fat body cell line (LC₅₀ = 7.55 μg/mL) was about twice more susceptible to Cry2Ab than the midgut cell line (16.0 μg/mL), the susceptibility of which was not significantly greater than that of SF9 cells (27.0 μg/mL). Further, ligand blot showed the binding differences between Cry1Ac and Cry2Ab in the three cell lines. These results indicated that the receptors of Cry2Ab were enriched in fat body cells and thus largely different from the receptors of Cry1Ac, which were enriched in midgut cells.

The full-length Cry1Ac protoxin without proteolytic activation exhibits toxicity against insect cell line CF-203

The new dual model for Bacillus thuringiensis insecticidal mechanism proposed that Cry1A protoxins without proteolytic activation could bind to insect midgut receptors to exert toxicity. To evaluate insecticidal potency of Cry1Ac protoxin at precluding interference of midgut proteases, the cytotoxicity of Cry1Ac protoxin against midgut cell line CF-203 derived from Choristoneura fumiferana was analyzed. It was revealed that Cry1Ac protoxin was toxic to CF-203 cells and there existed certain differences in the cytological changes when treated with protoxin and toxin. Our cell-based study provided direct evidence for the proposed dual model and shed light on exploring the difference between two toxic pathways elicited by intact protoxin and activated toxin.

Cell lines as models for the study of Cry toxins from Bacillus thuringiensis

Cell lines have been use extensively for the study of the mode of action of different pore forming toxins produced by different bacterial species. Bacillus thuringiensis Cry toxins are not the exception and their mechanism of action has been analyzed in different cell lines. Here we review the data obtained with different cell lines, including those that are naturally susceptible to the three domain Cry toxins (3d-Cry) and other non-susceptible cell lines that have been transformed with 3d-Cry toxin binding molecules cloned from the susceptible insects. The effects on Cry toxin action after expressing different insect gut proteins, such as glycosyl-phosphatidyl-inositol (GPI) anchored proteins (like alkaline phosphatase (ALP) aminopeptidase (APN)), or trans-membrane proteins (like cadherin (CAD) or ATP-binding cassette subfamily C member 2 (ABCC2) transporter) in cell lines showed that, with few exceptions, expression of GPI-anchored proteins do not correlated with increased susceptibility to the toxin, while the expression of CAD or ABCC2 proteins correlated with induced susceptibility to Cry toxins in the transformed cells lines. Also, that the co-expression of CAD and ABCC2 transporter induced a synergistic effect in the toxicity of 3d-Cry toxins. Overall the data show that in susceptible cell lines, the 3d-Cry toxins induce pore formation that correlates with toxicity. However, the intracellular responses remain controversial since it was shown that the same 3d-Cry toxin in different cell lines activated different responses such as adenylate cyclase-PKA death response or apoptosis. Parasporins are Cry toxins that are toxic to cancer cell lines that have structural similarities with the insecticidal Cry toxins. They belong to the 3d-Cry toxin or to MTX-like Cry toxin families but also show important differences with the insecticidal Cry proteins. Some parasporins are pore-forming toxins, and some activate apoptosis. In this review we summarized the results of the different studies about the Cry toxins mode of action using cultured cell lines and discuss their relation with the studies performed in insect larvae.

FOXA transcriptional factor modulates insect susceptibility to Bacillus thuringiensis Cry1Ac toxin by regulating the expression of toxin-receptor ABCC2 and ABCC3 genes

Cry toxins produced by Bacillus thuringiensis (Bt) are insecticidal proteins widely used in insect control. Recently, it was shown that ATP-binding cassette transporter proteins (ABC) such as ABCC2, ABCC3, ABCG1 and ABCA2 are implicated in the insecticidal action of Cry toxins as putative receptors. However, the transcriptional regulators involved in the expression of ABC transporter genes remain unknown. Sequence analysis of promoter regions of ABCC2 gene from Helicoverpa armigera and ABCC3 gene from Spodoptera litura Sl-HP cultured cells, revealed the potential participation of Forkhead box protein A (FOXA), a transcription factor that regulates the expression of genes through remodeling chromatin. To determine if FOXA was involved in regulating expression of ABCC2 and ABCC3 genes, the expression of FOXA, ABCC2 and ABCC3 was compared in Sl-HP cells that are sensitive to Cry1Ac toxin with those in S. frugiperda Sf9 cells that are not sensitive to the toxin. Expression levels of those genes were significantly higher in Sl-HP than in Sf9 cells. Transient expression of FOXA in Sf9 cells activated ABCC2 and ABCC3 transcription, which directly correlated with enhanced Cry1Ac-susceptibility in these cells. Silencing of FOXA gene expression by RNAi in H. armigera larvae resulted in a decreased expression of ABCC2 and ABCC3 without affecting expression of other Cry toxin receptor genes such as alkaline phosphatase, aminopeptidase or cadherin. Silencing of FOXA gene expression also resulted in a Cry1Ac-tolerant phenotype since lower mortality and higher pupation rate were observed in diet containing Cry1Ac protoxin in comparison with the control group. These results demonstrate that FOXA up-regulates expression of the Cry1Ac-toxin receptor ABCC2 and ABCC3 genes, and that lower FOXA expression correlates with tolerance to Cry toxin in cell lines and in lepidopteran larvae.

Toxicity of Cry1A toxins from Bacillus thuringiensis to CF1 cells does not involve activation of adenylate cyclase/PKA signaling pathway

Bacillus thuringiensis (Bt) bacteria produce Cry toxins that are able to kill insect pests. Different models explaining the mode of action of these toxins have been proposed. The pore formation model proposes that the toxin creates pores in the membrane of the larval midgut cells after interaction with different receptors such as cadherin, aminopeptidase N and alkaline phosphatase and that this pore formation activity is responsible for the toxicity of these proteins. The alternative model proposes that interaction with cadherin receptor triggers an intracellular cascade response involving protein G, adenylate cyclase (AC) and protein kinase A (PKA). In addition, it was shown that Cry toxins induce a defense response in the larvae involving the activation of mitogen-activated kinases such as MAPK p38 in different insect orders. Here we analyzed the mechanism of action of Cry1Ab and Cry1Ac toxins and a collection of mutants from these toxins in the insect cell line CF1 from Choristoneura fumiferana, that is naturally sensitive to these toxins. Our results show that both toxins induced permeability of K⁺ ions into the cells. The initial response after intoxication with Cry1Ab and Cry1Ac toxins involves the activation of a defense response that involves the phosphorylation of MAPK p38. Analysis of activation of PKA and AC activities indicated that the signal transduction involving PKA, AC and cAMP was not activated during Cry1Ab or Cry1Ac intoxication. In contrast we show that Cry1Ab and Cry1Ac activate apoptosis. These data indicate that Cry toxins can induce an apoptotic death response not related with AC/PKA activation. Since Cry1Ab and Cry1Ac toxins affected K⁺ ion permeability into the cells, and that mutant toxins affected in pore formation are not toxic to CF1, we propose that pore formation activity of the toxins is responsible of triggering cell death response in CF1cells.

Loop replacements with gut-binding peptides in Cry1Ab domain II enhanced toxicity against the brown planthopper, Nilaparvata lugens (Stål)

Bacillus thuringiensis (Bt) Cry toxins have been used widely in pest managements. However, Cry toxins are not effective against sap-sucking insects (Hemiptera), which limits the application of Bt for pest management. In order to extend the insecticidal spectrum of Bt toxins to the rice brown planthopper (BPH), Nilaparvata lugens, we modified Cry1Ab putative receptor binding domains with selected BPH gut-binding peptides (GBPs). Three surface exposed loops in the domain II of Cry1Ab were replaced with two GBPs (P2S and P1Z) respectively. Bioassay results showed that toxicity of modified toxin L2-P2S increased significantly (~9 folds) against BPH nymphs. In addition, damage of midgut cells was observed from the nymphs fed with L2-P2S. Our results indicate that modifying Cry toxins based on the toxin-gut interactions can broaden the insecticidal spectrum of Bt toxin. This method provides another approach for the development of transgenic crops with novel insecticidal activity against hemipteran insects and insect populations resistant to current Bt transgenic crops.

Anopheles gambiae Ag55 cell line as a model for Lysinibacillus sphaericus Bin toxin action

Binary toxin (Bin) produced by Lysinibacillus sphaericus is toxic to Culex and Anopheles mosquito larvae. It has been used world-wide for control of mosquitoes that vector disease. The Bin toxin interacts with the glucosidase receptor, Cpm1, in Culex and its orthologue, Agm3, in Anopheles mosquitoes. However, the exact mechanism of its mode of action is not clearly understood. It is essential to understand mode of action of Bin toxin to circumvent the resistance that develops over generations of exposure. A suitable model cell line will facilitate investigations of the molecular action of Bin toxin. Here we report Bin toxin activity on Ag55 cell line that has been derived from an actual target, Anopheles gambiae larvae. The Bin toxin, both in pro and active forms, kills the Ag55 cells within 24h. Bin toxin internalizes in Ag55 cells and also induces vacuolation as tracked by Lysotracker dye. The dose response studies showed that 1.5nM of Bin toxin is sufficient to induce vacuolation and Ag55 cell death. Presence of α-glucosidase gene (Agm3) expression in the Ag55 cells was also confirmed. Thus, Ag55 cells constitute an appropriate model system to decipher the mode of Bin action in mosquito larvae.

Two specific membrane-bound aminopeptidase N isoforms from Aedes aegypti larvae serve as functional receptors for the Bacillus thuringiensis Cry4Ba toxin implicating counterpart specificity

The interaction between Bacillus thuringiensis Cry toxins and their receptors on midgut cells of susceptible insect larvae is the critical determinant in toxin specificity. Besides GPI-linked alkaline phosphatase in Aedes aegypti mosquito-larval midguts, membrane-bound aminopeptidase N (AaeAPN) is widely thought to serve as a Cry4Ba receptor. Here, two full-length AaeAPN isoforms, AaeAPN2778 and AaeAPN2783, predicted to be GPI-linked were cloned and successfully expressed in Spodoptera frugiperda (Sf9) cells as 112- and 107-kDa membrane-bound proteins, respectively. In the cytotoxicity assay, Sf9 cells expressing each of the two AaeAPN isoforms showed increased sensitivity to the Cry4Ba mosquito-active toxin. Double immunolocalization revealed specific binding of Cry4Ba to each individual AaeAPN expressed on the cell membrane surface. Sequence analysis and homology-based modeling placed these two AaeAPNs to the M1 aminopeptidase family as they showed similar four-domain structures, with the most conserved domain II being the catalytic component. Additionally, the most variable domain IV containing negatively charged surface patches observed only in dipteran APNs could be involved in insect specificity. Overall results demonstrated that these two membrane-bound APN isoforms were responsible for mediating Cry4Ba toxicity against AaeAPN-expressed Sf9 cells, suggesting their important role as functional receptors for the toxin counterpart in A. aegypti mosquito larvae.

•

Two novel Aedes GPI-APN isoforms were functionally expressed in Sf9 cells.

•

Cells expressing each AaeAPN were more sensitive to Cry4Ba toxin cytolysis.

•

Specific binding of Cry4Ba to individual AaeAPN-expressing Sf9 cells was demonstrated.

•

These two AaeAPNs mediating Cry4Ba cytotoxicity serve as receptors in Aedes larvae.

•

AaeAPN models reveal four-domain organization with implications for toxin-counterpart specificity.