Transient exposure to low levels of insecticide affects metabolic networks of honeybee larvae

Abstract

The survival of a species depends on its capacity to adjust to changing environmental conditions, and new stressors. Such new, anthropogenic stressors include the neonicotinoid class of crop-protecting agents, which have been implicated in the population declines of pollinating insects, including honeybees (Apis mellifera). The low-dose effects of these compounds on larval development and physiological responses have remained largely unknown. Over a period of 15 days, we provided syrup tainted with low levels (2 µg/L(-1)) of the neonicotinoid insecticide imidacloprid to beehives located in the field. We measured transcript levels by RNA sequencing and established lipid profiles using liquid chromatography coupled with mass spectrometry from worker-bee larvae of imidacloprid-exposed (IE) and unexposed, control (C) hives. Within a catalogue of 300 differentially expressed transcripts in larvae from IE hives, we detect significant enrichment of genes functioning in lipid-carbohydrate-mitochondrial metabolic networks. Myc-involved transcriptional response to exposure of this neonicotinoid is indicated by overrepresentation of E-box elements in the promoter regions of genes with altered expression. RNA levels for a cluster of genes encoding detoxifying P450 enzymes are elevated, with coordinated downregulation of genes in glycolytic and sugar-metabolising pathways. Expression of the environmentally responsive Hsp90 gene is also reduced, suggesting diminished buffering and stability of the developmental program. The multifaceted, physiological response described here may be of importance to our general understanding of pollinator health. Muscles, for instance, work at high glycolytic rates and flight performance could be impacted should low levels of this evolutionarily novel stressor likewise induce downregulation of energy metabolising genes in adult pollinators.

Figure 1. Functional annotation of differentially expressed genes.

(A) Based on RNA-Seq data, expression levels for 300 genes were found to be significantly changed in imidacloprid-exposed larvae (compared with data obtained from non-exposed larvae). Expression is reduced for 195 genes (blue); expression is increased for 105 genes (red). The group of over-expressed genes includes nine cytochrome P450s; their gene IDs and normalised fold-change values are shown in the table. (B) Selection of non-redundant Gene Ontology (GO) terms for the three ontologies: Biological Process (BP), Molecular Function (MF) and Cellular Component (CC), and Interpro (IP) protein domains that are overrepresented in the 105 up (red) and 195 down (blue) regulated genes. Significance of enrichment is based on Fisher’s Exact Test. GO terms for the most significant enrichment groups are indicated in bold letters.

Figure 2. Expression of genes encoding carbohydrate-metabolising enzymes is affected in imidacloprid-exposed worker bee larvae.

Genes/enzymes, including paralogues, and their positions (coloured/grey boxes) in the glycolytic and related carbohydrate pathways are placed with reference to honeybee-specific pathway variations , . Based on DEGseq analysis, ten genes are downregulated (blue), including PGI, PGK, PGLYM, ENO and PYK, of the glycolytic pathway; PEPCK is upregulated (red). A key for gene names is provided; fold-changes in expression as determined by RNA-Seq (IE relative to C data set).

Figure 3. Multiplex RT-PCR validation of HSP90 and a P450 gene (CYP9-clade).

Variance in measured RNA expression levels was assessed between treatment groups (left), among hives (middle) and between the two measurements taken for individual larvae (right/larval samples 1–34 came from control hives C1–C3; Larval samples 35–74 came from imidacloprid-exposed hives IE1–IE3. As two measurements were made for each larva, the height of each bar in the larva plot gives the range of values recorded; the mean value for each larva is indicated in the middle of this range. For those larvae that yielded only a single successful measurement for a given gene, only one point is shown. Box and whisker plots show 25th percentile (bottom of box), 50th percentile (middle of box), 75th percentile (top of box), median (line in middle of box), and full range of data (low bar to high bar). Relative expression levels are compared to a standardisation factor. The scale of the y-axis differs between plots for the two genes and indicates fold-difference for the expression level relative to the standardisation factor.

Figure 4. Exemplary data of high resolution LC-MS analysis of larval lipid extracts.

Shown are graphic representations of summed spectra (positive electrospray mode) from a larval sample of an imidacloprid-exposed hive (A) and a larval sample of an unexposed, control hive (B). Y-axes (ion intensity) are normalised to the most intense ion species; x-axes indicate the specific mass-to-charge ratio (m/z). Smaller insets in A and B are zoomed-in portions of the graph axis scale (m/z 840–855) and provide an example of a typical change at m/z 850.587, which is elevated in imidacloprid-exposed samples. While some differences between the spectra can be readily observed, data processing enabled more detailed analysis.