Unravelling the Molecular Determinants of Bee Sensitivity to Neonicotinoid Insecticides

Abstract

The impact of neonicotinoid insecticides on the health of bee pollinators is a topic of intensive research and considerable current debate [1]. As insecticides, certain neonicotinoids, i.e., N-nitroguanidine compounds such as imidacloprid and thiamethoxam, are as intrinsically toxic to bees as to the insect pests they target. However, this is not the case for all neonicotinoids, with honeybees orders of magnitude less sensitive to N-cyanoamidine compounds such as thiacloprid [2]. Although previous work has suggested that this is due to rapid metabolism of these compounds [2-5], the specific gene(s) or enzyme(s) involved remain unknown. Here, we show that the sensitivity of the two most economically important bee species to neonicotinoids is determined by cytochrome P450s of the CYP9Q subfamily. Radioligand binding and inhibitor assays showed that variation in honeybee sensitivity to N-nitroguanidine and N-cyanoamidine neonicotinoids does not reside in differences in their affinity for the receptor but rather in divergent metabolism by P450s. Functional expression of the entire CYP3 clade of P450s from honeybees identified a single P450, CYP9Q3, that metabolizes thiacloprid with high efficiency but has little activity against imidacloprid. We demonstrate that bumble bees also exhibit profound differences in their sensitivity to different neonicotinoids, and we identify CYP9Q4 as a functional ortholog of honeybee CYP9Q3 and a key metabolic determinant of neonicotinoid sensitivity in this species. Our results demonstrate that bee pollinators are equipped with biochemical defense systems that define their sensitivity to insecticides and this knowledge can be leveraged to safeguard bee health.

Keywords: CYP9Q; P450; acetamiprid; bumble bee; honeybee; imidacloprid; neonicotinoids; pesticide; thiacloprid.

Figure 1

Toxicodynamics and Pharmacokinetics of Neonicotinoid Sensitivity in Two Bee Species (A) LD₅₀ values for imidacloprid and thiacloprid upon oral and topical application in A. mellifera and B. terrestris. Sensitivity thresholds are depicted according to EPA toxicity ratings [8]. Data for A. mellifera is taken from [9, 10], data for B. terrestris was generated in this study. Error bars display 95% CLs (n = 4). (B) Specific binding of thiacloprid and imidacloprid to both A. mellifera and B. terrestris nAChRs. Error bars display standard deviation (n = 3). (C) Sensitivity of A-p-methoxy-. mellifera to imidacloprid and thiacloprid before and after pretreatment with the insecticide synergist ABT (aminobenzotriazole). Error bars display 95% CLs (n = 3). See also Table S1.

Figure 2

Identification of Neonicotinoid Metabolising P450s in Honeybee and Bumble Bee (A and C) (A) Thiacloprid and imidacloprid hydroxylation by recombinantly expressed P450s of the A. mellifera CYP3 clade and (C) the CYP9 family in B. terrestris. The production of the hydroxy metabolite of each insecticide is displayed per mg of P450 protein (NS, not significant; ^∗∗Pc < c0.01, ^∗∗∗Pc < c0.001; Welch’s t test). Error bars display standard deviation (n = 3). (B) Phylogenetic tree with branch bootstrap values for A. mellifera (green) and B. terrestris (blue) P450 genes. Genes are grouped according to their adscription to different P450 clades. Branches within the CYP3 clade marked with a red dot indicate the position of A. mellifera CYP9Qs and their closest B. terrestris orthologs involved in thiacloprid metabolism, as shown in (A), (C), and (D). (D) Resistance ratio (RR) of transgenic Drosophila strains expressing A. mellifera AmCYP9Q1–3 or B. terrestris BtCYP9Q4-5 transgenes to thiacloprid and imidacloprid compared to a control line (flies of the same genetic background but without the transgene). Significance is referenced against this control line and based on non-overlapping 95% fiducial limits of LC₅₀ values (n = 3). See also Figures S1, S2, and S3.

Figure 3

Metabolism of Acetamiprid and Model Substrates by Honeybee and Bumble Bee CYP9Q Subfamily P450s (A and C) Acetamiprid N-desmethylation by recombinantly expressed CYP9Q1–3 of A. mellifera and (C) CYP9Q4–5 of B. terrestris. The production of N-desmethylated acetamiprid is displayed per mg of protein. Error bars display standard deviation (n = 3). (B and D) (B) Activity of CYP9Q1–3 and (D) CYP9Q4–5 against different fluorescent coumarin model substrates. Error bars display standard deviation (n = 3). Abbreviations: MC, 7-methoxycoumarin; MFC, 7-methoxy-4-trifluoromethyl coumarin; EC, 7-ethoxy coumarin; BFC, 7-benzyloxy-4-trifluoromethyl coumarin; EFC, 7-ethoxy-4-trifluoromethyl coumarin; BOMFC, 7-benzyloxymethoxy-4-trifluoromethyl coumarin; MOBFC, 7-p-methoxy-benzyloxy-4-trifluoro coumarin.

Figure 4

Tissue-Specific Expression and Functional Characterization of Honeybee and Bumble Bee Neonicotinoid-Metabolizing P450s (A) Relative expression (fold change) of A. mellifera and B. terrestris thiacloprid metabolising CYP9Q genes in different tissues of worker bees measured by qPCR. Significant differences (p < 0.01) in expression between tissues is denoted using letters above bars as determined by One-Way ANOVA with post hoc testing (Benjamini and Hochberg). (B and C) (B) Whole-mount in situ hybridization showing the distribution and abundance of the AmCYP9Q3 transcript in the brain of a worker bee in different neuronal cells and in (C) the Malpighian tubules and distal midgut. (D and E) Expression of green fluorescent protein in the Malpighian tubules and specific neurons of the Drosophila brain driven by the Malp-tub GAL4 line. (F) Sensitivity of transgenic Drosophila to thiacloprid when the Malp-tub GAL4 line is used to drive expression of AmCYP9Q3. Error bars display 95% CLs.