Functional redundancy in the toxic pathway of Bt protein Cry1Ab, but not Cry1Fa, against the Asian corn borer

Abstract

Crops genetically engineered to produce insecticidal proteins from the bacterium Bacillus thuringiensis (Bt) have been used extensively to control some major crop pests, but their benefits decrease when pests evolve resistance. Better understanding of the genetic basis of resistance is needed to effectively monitor, manage, and counter pest resistance to Bt crops. Resistance to Bt proteins in at least 11 species of Lepidoptera, including many important crop pests, is associated with naturally occurring mutations that disrupt one or more of three larval midgut proteins: cadherin and ATP-binding cassette proteins ABCC2 and ABCC3. Here, we determined how CRISPR/Cas9-mediated mutations disrupting cadherin, ABCC2, and ABCC3 singly and in pairs affect resistance to Bt proteins Cry1Ab and Cry1Fa in the Asian corn borer (Ostrinia furnacalis), which is the most damaging pest of corn in Asia and is closely related to the European corn borer (Ostrinia nubilalis), a major pest in Europe and North America. The results from bioassays of six knockout strains and their parent susceptible strain support a model in which Cry1Ab can kill larvae via one path requiring ABCC2 or another path requiring cadherin and ABCC3, whereas Cry1Fa uses only the first path. The model's predictions are generally supported by results from genetic linkage analyses and responses to Cry1Ab and Cry1Fa of Sf9 cells and Xenopus oocytes modified to produce cadherin, ABCC2, and ABCC3 singly or in pairs. The functional redundancy identified here for Cry1Ab could sustain its efficacy against O. furnacalis and may exemplify a widespread natural strategy for delaying resistance.

Keywords: Bacillus thuringiensis; Ostrinia; gene editing; resistance management; sustainability.

Fig. 1.

CRISPR/Cas9-mediated knockout of the genes OfABCC2, OfABCC3, or both on chromosome 15 of O. furnacalis. (A) Genomic structure of OfABCC2 and OfABCC3, each with 25 exons (blue and green boxes, respectively). Red lines and scissors show sgRNA target sites for knockouts: C2-sgRNA1 and C2-sgRNA2 for OfABCC2 (C2–KO), C3-sgRNA1 and C3-sgRNA2 for OfABCC3 (C3–KO), and C2-sgRNA1 and C3-sgRNA3 for both (C2+C3–KO). (B) Sequences for each knockout: wild-type (Top) and representative histograms from direct sequencing of PCR products after knockout (Bottom). In the wild-type sequences, the red letters show the protospacer adjacent motif (PAM) and the blue letters show the protospacer sequence. The paired dotted lines indicate where each deletion starts and ends in the wild-type sequence and where the deletion occurs in the knockout sequence.

Fig. 2.

Effects of knocking out the genes encoding cadherin (Cad), ABCC2 (C2), and ABCC3 (C3) singly and in pairs on toxicity of Bt proteins (A) Cry1Ab and (B) Cry1Fa to O. furnacalis. The resistance ratio (RR) is the LC₅₀ for a knockout strain divided by the LC₅₀ of the same toxin for the susceptible strain NJ-S (0.086 for Cry1Ab and 0.099 for Cry1Fa, both in μg toxin per g diet; (SI Appendix, Table S3). The black bars show the 95% fiducial limits of each RR. For the C2+C3–KO strain versus Cry1Fa, mortality was 19% at the highest concentration tested (750 μg Cry1Fa per g diet) and we could not accurately calculate the LC₅₀ or RR. (C) Model of toxic paths and effects of knockouts based on the bioassay results in (A) and (B). Both toxic paths are effective for Cry1Ab but only path 1 is effective for Cry1Fa. Strikethrough indicates knockout. S means RR < 2, R means RR > 4,000.

Fig. 3.

Toxicity of (A) Cry1Ab and (B) Cry1Fa to Sf9 cells infected to produce cadherin (Cad), ABCC2 (C2), or ABCC3 (C3) singly or in pairs. The black bars show the 95% fiducial limits for each LC₅₀ value. For Cry1Fa, cell death for the three types of cells that did not produce ABCC2 ranged from 12 to 15% at the highest concentration tested (30 nM) and we could not accurately calculate LC₅₀. At the highest concentration tested for each toxin, the mortality of control Sf9 cells (infected with the pFastBac Dual™ empty vector) was 7% for Cry1Ab at 2,000 nM and 11% for Cry1Fa at 30 nM.

Fig. 4.

Pore formation as indicated by the negative current caused by (A) Cry1Ab and (B) Cry1Fa in Xenopus oocytes transfected to produce cadherin (Cad), ABCC2 (C2), or ABCC3 (C3) singly or in pairs. Shown are means ± SEM based on recordings from six individual oocytes from two to three frogs. For the controls (untransfected oocytes injected with water) no negative current was caused by Cry1Ab (8.8 nA) or Cry1Fa (11.6 nA). Current was recorded 300 s after exposure to Cry1Ab or Cry1Fa.