Engineering Bacillus thuringiensis Cyt1Aa toxin specificity from dipteran to lepidopteran toxicity

Abstract

The Cyt and Cry toxins are different pore-forming proteins produced by Bacillus thuringiensis bacteria, and used in insect-pests control. Cry-toxins have a complex mechanism involving interaction with several proteins in the insect gut such as aminopeptidase N (APN), alkaline phosphatase (ALP) and cadherin (CAD). It was shown that the loop regions of domain II of Cry toxins participate in receptor binding. Cyt-toxins are dipteran specific and interact with membrane lipids. We show that Cry1Ab domain II loop3 is involved in binding to APN, ALP and CAD receptors since point mutation Cry1Ab-G439D affected binding to these proteins. We hypothesized that construction of Cyt1A-hybrid proteins providing a binding site that recognizes gut proteins in lepidopteran larvae could result in improved Cyt1Aa toxin toward lepidopteran larvae. We constructed hybrid Cyt1Aa-loop3 proteins with increased binding interaction to Manduca sexta receptors and increased toxicity against two Lepidopteran pests, M. sexta and Plutella xylostella. The hybrid Cyt1Aa-loop3 proteins were severely affected in mosquitocidal activity and showed partial hemolytic activity but retained their capacity to synergize Cry11Aa toxicity against mosquitos. Our data show that insect specificity of Cyt1Aa toxin can be modified by introduction of loop regions from another non-related toxin with different insect specificity.

Figure 1

Analysis of binding of Cry1Ab-G439D mutant toxin in loop3 of domain II to purified APN1, ALP or CAD proteins. ELISA plates were coated with 0.5 μg of each recombinant receptor protein (APN1, ALP or CAD) and the binding of different concentrations of wild type or mutant toxin was analyzed. Each experiment was performed in duplicate with a total of six repetitions for each toxin.

Figure 2

Location of loops regions in Cyt1Aa toxin where loop3 from Cry1Ab was inserted. Panel A, Amino acid sequence of Cyt1Aa toxin is aligned with the three-dimensional structures, helices alpha and beta strands are labeled over the amino acid sequence. The selected loop regions are labeled in red, arrows indicate the site of insertion of loop3 (FRSGFSNSSVSI) from Cry1Ab. Residues labeled in green correspond to regions involved in synergism with Cry11Aa and Cry4Ba. Panel B, Localization of loops in the three dimensional structure of Cyt1Aa (PDB 3RON) showing that the selected regions were exposed to the solvent. Loop regions that resulted in improved toxicity against lepidopteran larvae are in blue dark color. Loop regions important for synergism are in green color.

Figure 3

Purification of Cyt1Aa-hybrid proteins expressed in B. thuringiensis 407^- strain. Panel A, purified trypsin activated toxins resolved in SDS-PAGE 15% acrylamide stained with Coomassie blue. Panel B, analysis of hybrid Cyt1Aa protoxins by western blot assays using anti-Cyt1Aa polyclonal specific antibody. Panel C, analysis of hybrid Cyt1Aa activated toxins by western blot assays using anti-Cyt1Aa polyclonal specific antibody. Page ruler plus pre-stained protein ladder in panels B and C are shown as separate images.

Figure 4

Analysis of binding of Cyt1Aa and hybrid-Cyt1Aa variants to purified APN1, ALP or CAD proteins. ELISA plates were coated with 0.5 μg of each recombinant receptor protein (APN1, ALP or CAD) and the binding of 100 nM Cyt1Aa or hybrid-Cyt1Aa was analyzed. Each experiment was performed in duplicate with a total of six repetitions for each mutant Cyt1Aa toxin. *Indicate significant differences with the Cyt1Aa toxin by ANOVA with significant differences P < 0.01.

Figure 5

Analysis of synergism of Cyt1Aa or Cyt1Aa-hybrid toxins with Cry11Aa toxin. Cry11Aa toxin was used at 108 ng/ml and Cyt1Aa or hybrid-Cyt1Aa were used at 30 ng/ml. Negative control (dechlorinated water) was included in the bioassay. Larvae mortality was examined 24 h after treatment. *Indicate significant differences with the Cyt1Aa toxin by ANOVA with significant differences P < 0.01.

Figure 6

Hemolytic activity of Cyt1Aa and hybrid-Cyt1Aa mutants, Hemolysis was analyzed with rabbit red blood cells. These assays were performed three times and standard deviations are shown.

Figure 7

Toxicity of Cyt1Aa or hybrid-Cyt1Aa variants to Manduca sexta or Plutella xylostella larvae. M. sexta neonate or P. xylostella third instar larvae were treated with 2 μg of Cyt1Aa or hybrid-Cyt1Aa per cm² artificial diet. The mortality was monitored after 15 days. Twenty-four larvae were used per toxin concentration in triplicate. Letters above the bars indicate significant differences with the Cyt1Aa toxin by ANOVA with significant differences P < 0.01.