Resistance mutation conserved between insects and mites unravels the benzoylurea insecticide mode of action on chitin biosynthesis

Abstract

Despite the major role of chitin biosynthesis inhibitors such as benzoylureas (BPUs) in the control of pests in agricultural and public health for almost four decades, their molecular mode of action (MoA) has in most cases remained elusive. BPUs interfere with chitin biosynthesis and were thought to interact with sulfonylurea receptors that mediate chitin vesicle transport. Here, we uncover a mutation (I1042M) in the chitin synthase 1 (CHS1) gene of BPU-resistant Plutella xylostella at the same position as the I1017F mutation reported in spider mites that confers etoxazole resistance. Using a genome-editing CRISPR/Cas9 approach coupled with homology-directed repair (HDR) in Drosophila melanogaster, we introduced both substitutions (I1056M/F) in the corresponding fly CHS1 gene (kkv). Homozygous lines bearing either of these mutations were highly resistant to etoxazole and all tested BPUs, as well as buprofezin-an important hemipteran chitin biosynthesis inhibitor. This provides compelling evidence that BPUs, etoxazole, and buprofezin share in fact the same molecular MoA and directly interact with CHS. This finding has immediate effects on resistance management strategies of major agricultural pests but also on mosquito vectors of serious human diseases such as Dengue and Zika, as diflubenzuron, the standard BPU, is one of the few effective larvicides in use. The study elaborates on how genome editing can directly, rapidly, and convincingly elucidate the MoA of bioactive molecules, especially when target sites are complex and hard to reconstitute in vitro.

Keywords: CRISPR/Cas9; benzoylureas; insecticide resistance; mosquito control; resistance management.

Fig. S1.

Chemical structures of insecticides used in this study. According to the IRAC classification system (www.irac-online.org/modes-of-action/), Etoxazole belongs to group 10B (mite growth inhibitors), BPUs (Diflubenzuron, Lufenuron, Triflumuron) belong to group 15 (inhibitors of chitin biosynthesis, type 0), whereas Buprofezin belongs to group 16 (inhibitors of chitin biosynthesis, type 1). Cyromazine belongs to group 17 (moulting disruptor, dipteran).

Fig. 1.

Log-dose mortality data for triflumuron tested against third instar larvae of diamondback moth strains BCS-S, Sudlon, and Sudlon-Tfm as well as combined reciprocal crosses (F1). Error bars represent SEM.

Fig. S2.

Comparison of the postembryonic developmental time (±SD) of strains Sudlon and Sudlon-Tfm. Sudlon-Tfm shows a significant longer larval (L4) and pupal development (P < 0.0005, R² = 0.3, n = 50).

Fig. 2.

Location of the two mutations conferring resistance. (Top) Schematic representation of domain architecture of CHS1, redrafted from ref. . 5TMS, cluster of five transmembrane segments; CC, coiled-coil motif; CD, catalytic domain; CTR, C-terminal region; NTR, N-terminal region. Rectangular boxes represent transmembrane domains. Arrows point to signature sequences QRRRW (catalytic domain) and WGTR (N-terminal region). (Bottom) Aligned amino acid sequences of helix 5 in the 5TMS clusters of CHS1 of D. melanogaster (Dm), M. sexta (Ms), six strains of P. xylostella (Px), and T. urticae (Tu; S, etoxazole susceptible; R, etoxazole resistant). Conserved residues are shown in bold. The position of the I1042M/F substitution in resistant P. xylostella (I1017F in etoxazole-resistant mites) is indicated in gray.

Fig. S3.

SNP pyrosequencing assay results for the CHS I1042M mutation (ATT/ATG, printed in bold letters) in amplified cDNA and gDNA fragments of P. xylostella. (A) Homozygous ATT, genotype SS. (B) Heterozygous ATT/ATG, genotype RS. (C) Homozygous ATG, genotype RR. (D) Heterozygous TTT/ATT, genotype RS. (E) Heterozygous TTT/ATG, genotype RR. (F) Homozygous TTT, genotype RR.

Fig. S4.

Phylogenetic analysis of several arthropod CHS protein sequences. Dipteran CHSs: A. gambiae (accession no. CH1: XP_321336.5, CH2: XP_321951.2), Bactrocera dorsalis (CHS1: XP_011203784.1, CHS2: AGC38392.1), Ceratitis capitata (CHS1: XP_012157009.1, CHS2: XP_012161954.1), Drosophila grimshawi (CHS1: XP_001994028.1, CHS2: XP_001985562.1), D. melanogaster (kkv: NP_524233.1, CHS2: NP_524209.3). Other insect CHSs: P. xylostella (Lepidoptera, BAF47974.1), Helicoverpa armigera (Lepidoptera, AKJ54482.1), Tribolium castaneum (Coleoptera, NP_001034491.1), Anasa tristis (Hemiptera, AFM38193.1). Other arthropod CHSs include spider mite T. urticae (XP_015781017.1), horseshoe crab Limulus polyphemus (XP_013790798.1), and crustacean Daphnia magna (KZS08010.1). Phylogenetic analysis was performed as described in ref. using the one-click mode (www.phylogeny.fr/index.cgi). Numbers on branches indicate the approximate likelihood ratio test confidence index.

Fig. S5.

Nucleotide and deduced amino acid sequence of an 800-bp fragment of kkv exon 6 (corresponding to 3R:5381406:5382206 at the BDGP6 genome assembly), flanking position 1056 of the D. melanogaster amino acid sequence (I, shown in black background), equivalent to 1042 in Plutela xylostella and 1017 in T. urticae. Light gray areas indicate the CRISPR/Cas9 targets selected (gRNA444, gRNA658), whereas dark gray areas indicate the corresponding PAM (–NGG) triplets. Vertical arrows denote break points for CRISPR/Cas9-induced double stranded breaks. Ovals mark differences between target (wild-type) and donor (genome modified) sequences (red for I1056M and blue for I1056F). A C→T synonymous transition common for both designs at position 407 abolishes an NcoI cleavage site (CCATGG, underlined); an A→T transversion at position 492 generates a codon alteration (ATC→TTC) that results in the I1056F mutation, whereas a C→G transversion at position 494 generates a different codon alteration (ATC→ATG) that results in the I1056M mutation. Six extra synonymous mutations present at the CRISPR targets in the I1056M design are shown in red letters. Horizontal arrows indicate the relative positions of the primers used for diagnostic screening [ETXSF, ETXSR, ETXSM (blue, specific for I1056F), and PxM (red, specific for I1056M); see Table S2].

Fig. S6.

Screening for genome-modified flies. (Top) PCR screening after digestion with NcoI of template DNA from pools of G₁ flies derived from different G₀ (injected) individuals using a specific primer pair (ETXS/PxM) for I1056M mutation [–, nos.Cas9 DNA (negative control); +, donor plasmid template (positive control)]. (Bottom) Screening of G₂ flies from line Et15 crossed to balancer TM3Sb, PCR amplification with a generic primer pair (ETXSF/ETXSR), and digestion of 395-bp product with NcoI. The modified allele remains uncut, whereas the wild-type allele is cut in two smaller bands; positives are heterozygous at this stage.

Fig. S7.

(Top) Comparison of developmental timing and number of eclosed adults out of 50 original eggs among different Drosophila lines used in the bioassays of this study. No significant difference was found between Px39 (I1042M) or Et15 (I1017F) versus nos.Cas9 wild-type controls (P = 0.151 and P = 0.4, respectively). (Bottom) Comparison of average daily fecundity among lines (n = 8 for Et15 and nos.Cas9, n = 7 for Px39). No evidence for reduced fecundity is found for line Px39 (I1042M). Error bars represent SEM.