Article | Open | Published:

Chronic exposure to a neonicotinoid increases expression of antimicrobial peptide genes in the bumblebee Bombus impatiens

Scientific Reports volume 7, Article number: 44773 (2017) | Download Citation


Bumblebees are important pollinators in wild and agricultural settings. In recent decades pollinator declines have been linked to the effects of increased pesticide use and the spread of disease. Synergy between these factors has been suggested, but no physiological mechanism has been identified. This study examines the connection between neonicotinoid exposure and innate immune function in the bumblebee Bombus impatiens, which is an important wild and commercial pollinator in eastern North America. Experimental colonies in the field were enclosed and provided pollen and sugar syrup containing an agriculturally relevant range of imidacloprid concentrations. Bumblebees were collected from colonies over four weeks, and the expression of antimicrobial peptides was measured using multiplex quantitative real time PCR. Significant increases in the expression of abaecin, apidaecin and hymenoptaecin were found over time in treatments receiving moderate to high concentrations of the pesticide. Responses were dependent on time of exposure and dose. These results indicate that immune function in bumblebees is affected by neonicotinoid exposure and suggest a physiological mechanism by which neonicotinoids may impact the innate immune function of bumblebee pollinators in wild and agricultural habitats.


Pollination is a critical ecosystem service, essential for the persistence of wild habitats as well as human agriculture1. The rapid growth in agricultural demand means that managed pollinator species, such as honeybees, are unlikely to meet the demand for agricultural pollination on their own2, increasing the importance of pollination by wild insects3. Bumblebees play a key role in pollination, but are experiencing widespread declines in wild populations4,5. The cause of these declines has been the focus of much attention, and may in fact be multi-factorial6,7. Some of the potential influences that have been suggested include the spread of parasites as well as exposure to pesticides. In particular, neonicotinoid pesticides have been a focus of concern8. Because of the critical role of wild insect pollinators and the danger posed by their decline, greater understanding of their biological response to these factors is imperative.

Neonicotinoids have been shown to negatively affect bees in several ways, and collectively it is thought that these off-target effects limit the effectiveness of commercial and natural pollination services. Since they are systemic, neonicotinoids or their metabolites can be present in the nectar and pollen of plants after application, with the potential for chronic exposure to pollinators9,10,11,12,13,14. Neonicotinoids are neurotoxic and have been found to alter foraging behavior, reduce foraging efficiency15,16,17,18, and increase displacement of queens in honeybees19. Similar effects of neonicotinoid exposure have been reported in the European buff-tailed bumblebee Bombus terrestris20,21. Exposure has also been linked to reduced colony growth and reproduction of B. terrestris9,22,23,24. Individual survival of B. terrestris is also lower as a result of neonicotinoid treatment9,23,25,26. A recent study of the North American common eastern bumblebee Bombus impatiens found similar negative effects for two neonicotinoids on queen mortality and foraging efficiency27. Furthermore, pathogen challenge experiments using honeybees exposed to neonicotinoids resulted in increased levels of Nosema28 and deformed wing virus29. However, at low concentrations (1 ppb clothianidin), effects on learning, fecundity and disease susceptibility may vary30.

Despite previous work characterizing neonicotinoid effects on health, there has been relatively little investigation of mechanisms by which neonicotinoids might increase bees’ susceptibility to infection31. One possibility is that increased rates of infection following neonicotinoid exposure might be explained by suppression of individual insects’ innate immune system21 or of a colony’s social immunity. Several recent studies have addressed this possibility in honeybees. Alaux and colleagues6 found that honeybee mortality resulting from Nosema infection increased significantly with exposure to imidacloprid, a neonicotinoid commonly used in agriculture. In their study, phenoloxidase activity and hemocyte numbers, reflecting individual immunity, were not altered. However glucose oxidase, which functions in colony-level immunity, was significantly reduced in treatments receiving both Nosema and imidacloprid, suggesting a synergistic effect. Brandt and coworkers32 found that exposure to a similar concentration of imidacloprid, as well as to the neonicotinoids thiacloprid and clothianidin, did suppress individual immunity, as measured by hemocyte numbers, foreign-body encapsulation, and the antimicrobial activity of hemolymph after 24 hours. The authors suggest the differences in their results may stem from differing responses of healthy bees versus those with prior exposure to Nosema.

Antimicrobial peptides (AMPs) are a major component in insect innate immunity and exist in a diverse range of organisms33. The AMPs of insects have been considered most effective against Gram-positive bacteria, with more limited activity against Gram-negative bacteria and fungi34,35. Wounding and infection with E. coli cause increased expression of AMPs in B. terrestris36. Another recent study of this bumblebee species demonstrated that the AMPs abaecin and defensin were also up-regulated during infection with the trypanosome Crithidia bombi. Suppression of these genes with RNA interference also led to increased parasite titers37. Similarly, exposure of honeybees to Varroa mites increased AMP expression38, although this may have been due to secondary viral infection. AMP expression in honeybees is down-regulated after infection with Nosema ceranae, however this parasite is thought to suppress host immunity39.

Work with honeybee larvae has begun to explore potential influences of neonicotinoid exposure on AMP expression. Gregorc and colleagues38 examined the effects of a high imidacloprid concentration on the expression of 23 immunity-related genes, finding increased expression of prophenoloxidase-activating enzyme, but no significant effects on transcript levels for five AMP genes. That study also found no synergistic effects of pesticide exposure and mite parasitism of larvae. Di Prisco and coworkers29 examined the effects of multiple neonicotinoids on honeybees and reported that clothianidin treatment caused up-regulation of an inhibitor of NF-κB, as well as reduced expression of the AMP apidaecin. In both these studies, organophosphate insecticides did not produce comparable changes in gene expression. Aufauvre and coworkers40 examined the influences of imidacloprid exposure and Nosema ceranae infection on honeybee workers, finding reduced expression of hymenoptaecin, alone among AMPs, after exposure to both the pesticide and parasite.

No study to date has tested whether neonicotinoids can alter the expression of antimicrobial peptide genes in bumblebees. Therefore, the goal of this study was to test this hypothesis by examining the expression of four AMPs in the common eastern bumblebee Bombus impatiens. This was done over one month, following exposure to a range of agriculturally realistic concentrations of the neonicotinoid imidacloprid. Bombus impatiens is an increasingly important pollinator in US commercial agriculture41,42,43,44, and wild populations are common across eastern North America45. We provide evidence of variation in AMP expression in bumblebees in response to chemical exposure. In field-enclosed colonies, imidacloprid exposure significantly affected transcript levels for three of the four AMPs examined. Surprisingly, the direction of change was an increase in AMP expression following exposure.


Variation in AMP expression in different body regions

In order to determine the optimal sampling method, we examined AMP expression in different body regions of bees collected from separate colonies in field enclosures prior to pesticide treatment (Fig. 1A–D). Normalized AMP expression in the thorax and abdomen was similar in magnitude to samples from untreated whole foragers. However, expression of apidaecin was significantly lower in heads compared to abdomens (ANOVA F2,15 = 5.03, p = 0.0213; Tukey’s test head-abdomen p = 0.0197), with marginally less variance. Because of this difference, all other sampling was made using homogenates of whole adult bodies.

Figure 1
Figure 1

Expression of antimicrobial peptide genes encoding abaecin, apidaecin, defensin and hymenoptaecin from different body regions (AD) and developmental stages (EH) of Bombus impatiens raised in untreated, field enclosed colonies. The range of expression is indicated by Tukey’s plot: boxes demarcate the upper and lower quartiles, while the heavy bar indicates the median of normalized expression. Whiskers extend to 1.5 times the interquartile range or the most extreme value. Group differences according to post-hoc tests are indicated by different letters appearing above the whiskers for each group. Panels without letters indicate that no significant differences were found. Six individuals were used to examine expression by body region, and color-coding consistently indicates the same individual. Samples sizes for brood of each stage were 6 early larvae, 4 mid-stage larvae, 8 late larvae, and 5 pupae.

Developmental time course of AMP expression

Two AMPs, apidaecin and hymenoptaecin, did not vary in their normalized expression among brood of different stages (Fig. 1F and H). An increase in normalized expression over development was detected for abaecin (Fig. 1E), with the lowest expression in mid-stage larvae and significantly higher normalized expression in pupae (Kruskal-Wallis rank sum test χ23 = 12.129, p = 0.00695; Dunn’s test, pupae vs. mid-stage larvae p = 0.0026, late stage larvae p = 0.0355). In contrast, normalized expression of defensin (Fig. 1H) was highest in early larvae and decreased in older stages, with significantly lower expression in pupae (Kruskal-Wallis rank sum test χ23 = 9.2955, p = 0.0256; Dunn’s test, early larvae vs. pupae p = 0.0099). These results show that AMPs in B. impatiens are subject to independent regulation in some biological contexts.

AMP expression increased after imidacloprid exposure

Expression of four antimicrobial peptide genes was measured in bumblebees collected from colonies treated with a range of imidacloprid concentrations, administered in pollen and syrup. Concentrations were chosen to span the range of imidacloprid found in pollen and nectar of common crops after pesticide application (Table 1)10,11,46,47. In all experiments, hives that were not exposed to imidacloprid showed no significant change in mean AMP expression over time (Table 2). This includes colonies raised in the same location as previously treated colonies in the following year (Supplementary Fig. 1). However, significant correlations between the exposure time and the expression of AMPs were seen for all imidacloprid treatments (Table 2).

Table 1: Imidacloprid doses used in each treatment.
Table 2: Time-dependency of imidacloprid effects on AMP expression.

One possible confounding factor in this experiment is the possibility of infection. While it has not been demonstrate that infection induces AMP expression in B. impatiens, this has been shown in other insects39,48, including B. terrestris36. Examination of the abdominal interior did not reveal the presence of parasites in any of the sampled bees. Moreover, 18 individuals collected at day 19 from the control and high-concentration treatments were screened using realtime RT-PCR for the presence of Nosema bombi and Crithidia bombi, two of the most common pathogens of B. impatiens. No infections were identified (Supplementary Dataset S1).

Colonies in field enclosures that were treated with imidacloprid showed dose-dependent increases in normalized AMP expression over time (Fig. 2A,B and D). Significant dose-dependent positive correlations (Table 3) were found for the expression of abaecin after 9 days of exposure and for the duration of the experiment thereafter. Correlations to imidacloprid dose for expression of apidaecin and hymenoptaecin were also significant by day 31.

Figure 2: Expression of antimicrobial peptide genes from B. impatiens in captive colonies under differing imidacloprid dosage.
Figure 2

Panels A–D show log10 expression over time, normalized against actin-5C. Dots represent individual samples. For each treatment group, a linear regression to the mean for each time point is shown with the gray area indicating the 95% confidence interval. Panels E-H show changes in the normalized log10 expression of each gene, from the first to last time points (+/− standard error). Treatments are shaded by dosage, consistently in both the top and bottom panels. For abaecin, apidaecin and hymenoptaecin, the effect of dosage was significant according to the Kruskal-Wallis rank sum test. Different letters indicate significant pairwise differences identified by Dunn’s test with Bonferroni correction. Sample sizes for most treatments at each time point are 3 individuals from 3 colonies for 9 total individuals. Exceptions include low dose, time zero, with n = 8; low dose, day 19, n = 7; control, day 31, n = 6 (from 2 colonies); low dose, day 31, n = 6; medium and high doses, day 31, n = 5.

Table 3: Dose-dependency of AMP expression.

After 31 days of exposure to imidacloprid, expression differed significantly for abaecin, apidaecin and hymenoptaecin (Fig. 2; Kruskal-Wallis test for abaecin χ23 = 11.19, p = 0.0107; for apidaecin χ23 = 12.28, p = 0.00648; for hymenoptaecin χ23 = 8.300, p = 0.0402). However defensin expression in each of the treated groups was not significantly different from the control treatment (Fig. 2C and G). Colony-level change over time in the expression of abaecin and apidaecin was significantly greater for imidacloprid treatments, starting at the 12 μg/kg pollen dose (Fig. 2E,F; Dunn’s test abaecin control vs.12 μg/kg p = 0.0292, vs. 24 μg/kg p = 0.0159, apidaecin control vs.12 μg/kg p = 0.0028, vs. 24 μg/kg p = 0.0372). Expression change in hymenoptaecin was significantly greater than controls only in the highest concentration treatment, 24 μg/kg (Fig. 2H; Dunn’s test p = 0.0176). The final masses of treated and control colonies, including all bees, brood, comb, and stored products, were not significantly different among treatments.

A replicate of the experiment was conducted the following year, in the same locations, without imidacloprid exposure (Fig. S1). In this mock experiment, no significant changes over time were detected for abaecin, apidaecin and hymenoptaecin. However, two mock-treatment blocks has significant differences in their change in defensin expression over the course of the experiment (Kruskal-Wallis χ23 = 9.9038, p = 0.0194; Dunn’s test, blocks B–C p = 0.0325). Nevertheless, we interpret the outcome of this replicate to support the conclusion that changes in AMP expression in the first experiment resulted from imidacloprid exposure rather than confounding factors.

Expression of AMPs in wild B. impatiens

To compare the AMP expression levels in captive B. impatiens to those in wild populations (Fig. 3), we sampled foraging B. impatiens workers from pasture 3 km from the site of our field enclosures (Oakland, Maine), and at two botanical gardens, on the Atlantic coast (Boothbay, Maine) and inland (Boylston, Massachusetts). For all AMPs examined, a group consisting of all untreated bees from captive screen houses was not significantly different from wild B. impatiens collected at the two sites in Maine. However, bees collected from Boylston had significantly higher normalized expression of all AMPs than captive bees (Fig. 3). Expression of hymenoptaecin was also significantly higher in bees from Boylston compared to those collected from Boothbay. While Oakland was the closest site to the experimental screen houses, means of normalized expression for all AMPs were closer to values from Boothbay. It is unclear what local environmental factors might influence differences in AMP expression among collection sites. Expression of all 4 AMPs in wild B. impatiens was significantly correlated with the longitude of the collection site (Spearman’s rank correlation, abaecin ρ = −0.791, p = 0.0112; apidaecin ρ = −0.738, p = 0.0232; defensin ρ = −0.632, p = 0. 0.0676; hymenoptaecin ρ = −0.896, p = 0.00108). It is unclear what the proximate causes of this variation may be. However, it is possible that expression differences arise from variations in local pesticide exposure.

Figure 3: Expression of antimicrobial peptide genes in wild and captive B. impatiens, represented by Tukey’s plots.
Figure 3

Expression is shown on a log10 scale, normalized against actin-5c. Wild bees were collected from field sites in Boylston, Massachusetts (n = 3), Oakland, Maine (n = 3), and Boothbay, Maine (n = 3). Samples were also taken from untreated colonies in experimental screen houses (n = 161) in Waterville, Maine. For each gene, the Kruskal-Wallis rank sum test identified locality as a significant factor Different letters indicate significant pairwise differences as determined by Dunn’s test with Bonferroni correction.


Here we report the first evidence that neonicotinoid exposure can alter the expression of antimicrobial peptide genes in bumblebees. It has been suggested that neonicotinoids may contribute to pollinator declines by increasing susceptibility to disease6,7,31. However, no physiological mechanism for this ecological pattern has been identified. One potential mechanism for a synergy of neonicotinoids and bee disease is the suppression of genes functioning in immunity after exposure to pesticide29. Published data on antimicrobial peptide expression in honeybees after neonicotinoid exposure have been ambiguous. One study reported reduced abaecin expression after clothianidin treatment29, while another found hymenoptaecin, alone among AMPs, was reduced in a treatment including imidacloprid exposure and Nosema ceranae infection40. In contrast, a study by Gregorc and colleagues38 exposed honeybees to a high imidacloprid concentration, but detected no changes in AMP expression. Based on these reports, we predicted AMP expression in B. impatiens might be reduced after chronic exposure to imidacloprid.

Contrary to these predictions, imidacloprid exposure in this study was associated with increased expression for three of the four AMP-encoding genes that were measured in a time- and dose-dependent manner (Fig. 2; Tables 2, 3). While the strongest effects were seen in the highest concentration treatments, these concentrations are within field-realistic levels for some crops, such as squash46,47. There are several possible interpretations of these results.

An increase in AMP expression at the doses used in this study may represent the excitatory phase of a hormetic or biphasic reaction norm49,50. If this is true, then similar experiments using higher concentrations of imidacloprid are predicted to eventually lead to decreased expression of AMP-encoding genes. Our study used concentrations of imidacloprid in order to approximate the conditions of chronic exposure associated with crops having relatively high systemic imidacloprid46,47. A hormetic response of innate immunity to neonicotinoids would significantly complicate the interactions of pollinators, pathogens and pesticides.

Another possible explanation for increased AMP expression is as a general physiological response to stress induced by the pesticide. For example, in one study of immunity-related gene expression, honeybees injected with saline buffer had similar increases in the expression of AMPs and other genes as those challenged with E. coli or Paenibacillus larvae51. However, while that study found correlated changes in the transcript levels of all sampled AMPs as well as other genes, the fact that change was not observed for defensin in our study suggests that effects are more direct and not a generalized up-regulation of immunity. In the future, it will be possible to use high-throughput sequencing to examine pesticide-induced changes in gene expression more broadly.

A direct effect on AMP expression is also supported by the fact that hive masses were not significantly different after imidacloprid exposure in our experiments. A previous study of B. terrestris reported significantly less growth in colonies dosed with imidacloprid23. However, that experiment dosed colonies in lab before allowing bees to freely forage. Our experiments limited feeding to screen houses in order to provide consistent chronic exposure. Since neonicotinoid exposure is known to impair foraging effectiveness20,27, higher growth in captivity might be explained by lower physical and perceptual demands on bees obtaining food compared to those freely foraging.

Finally, imidacloprid-induced increase in AMP expression may be unique to this species of bumblebee. While many Bombus species around the world have undergone population declines4,7, B. impatiens populations in North America have remained stable and abundant45. Therefore, it is possible that this species is atypical in its response, compared to other native pollinators or to B. terrestris, a European native that has served as a model social insect and pollinator.

Our data demonstrate that neonicotinoids can be modulators of immune gene expression in B. impatiens and suggest that local environmental factors, such as pesticide exposure, may produce significant differences in immune activity of B. impatiens. In the northeast United States, B. impatiens and other native bumblebees are crucial wild pollinators of regional crops, such as blueberry41, cranberry42 and raspberry43. More ecotoxicological research with B. impatiens is needed to better understand how this species might fare in the future in environments that are wild, agricultural or otherwise synanthropic.


Bumblebee colonies

Colonies of Bombus impatiens were obtained from Koppert Biological Systems, Inc. (Howell, Michigan, USA). In May of 2014 and 2015, colonies were started with one queen and 30 workers, randomly divided between treatment groups. All colonies were housed in a 6-L plastic mesh cage, with a gated entrance, covered in cardboard. To provide experimental conditions approximating the conditions of wild bees, experimental colonies were allowed limited foraging. Colonies were enclosed in screen houses covering 20-m2 of previously mown marginal meadowland in Waterville, Maine, USA. Flowering plants were removed from within the enclosures to ensure that bees consumed the food provided during the experiment. Bees were provided with honeybee pollen (YS Eco Bee Farms, Sheridan, Illinois, USA) homogenized in a dry rotary homogenizer (Proctor Silex, Glen Allen, Virginia, USA) for at least 1 minute. Pollen and 60% (v/v) cane sugar syrup were available at wooden feeding stations roughly 1 m from hive entrances, at a height of 1 m. All food was replaced when empty or after rain. Native bumblebees, including B. impatiens, B. ternarius, and B. vagans, normally use the study area for foraging and nesting. However, no native colonies were allowed in the area during the study. Hive masses were measured upon completion of the experiment on a portable balance as the weight of all bees, brood, comb, honey, pollen and other hive products, as well as the weight of the plastic cage.

Experimental design

Imidacloprid of analytical grade (>98% purity; Sigma-Aldrich, St. Louis, Missouri, USA) was administered through pollen and syrup. Using distilled water, a 28 mg/L stock solution of imidacloprid was prepared using quantitative transfer. This stock was used to prepare syrup to the intended dosages. Pollen doses were prepared by combining powdered imidacloprid and homogenized pollen in a dry rotary homogenizer. Series dilutions (mass/mass) were used to obtain the treatment dosages.

Colonies were allowed to acclimate in situ for one week, before sampling or pesticide dosage. Starting May 2014, experimental colonies were distributed randomly among a vehicle control group and three imidacloprid dosage treatments. In each treatment, three colonies were provided untreated food or pollen and sugar syrup containing imidacloprid at low, medium or high doses (Table 1). These concentrations were chosen to reflect a realistic range that might be encountered by bees in the nectar and pollen on common crop plants10,11,46,47 and to correspond to similar, previous experiments of B. terrestris toxicology23.

In order to test whether trends observed in AMP expression were artifacts produced by the location of colonies in each treatment, the study design was replicated in 2015. Another set of colonies supplied by the same vendor was placed in the same locations and screen houses. All colonies received undosed food during this season, and samples were collected to examine variation over time.


Sampling was performed with colony- and individual-level replicates. Periodically, individuals were collected in triplicate from each colony. Collection targeted actively foraging bees. Therefore, individuals were collected emerging from or returning to nest boxes. Small emerging bees, which could potentially have been nurse bees, were avoided. Bees were collected during mid-day in 45-ml Corning snap-cap vials and anesthetized using FlyNap, a vapor of 50% triethylamine and 25% ethanol (Carolina Biological Supply Company, Burlington, North Carolina, USA). Individuals were then immediately decapitated, bisected and placed into RNAlater RNA Stabilization Reagent (Qiagen, Hilden, Germany). Within 1–4 hours, samples were transported to the laboratory and placed at −80 °C prior to RNA extraction.

In order to determine whether captive, commercially supplied bees were representative of wild B. impatiens, we also collected wild foraging B. impatiens from two sites in Maine and Massachusetts during June and July 2015. Mown, disused pastureland in Oakland, Maine, 3 km from the site of our field enclosures, was used as the most immediate comparison. Bumblebees were also collected from two botanical gardens, one on the Atlantic coast (Boothbay, Maine) and one roughly 50 km inland (Boylston, Massachusetts). At each site, three foraging B. impatiens was collected and stored in RNAlater as described above.

To determine an optimal sampling strategy for AMP gene expression measurement, additional bees collected from captive colonies were partitioned into head, thoracic (including legs and wings) and abdominal regions. Brood was also sampled from undosed colonies at the conclusion of field experiments in 2015. Brood was sorted by developmental stage and immediately used in RNA extraction.

Quantification of Gene Expression

Expression of AMPs was measured using two-step quantitative real time PCR with multiplexed dual-labeled probes. Tissues were moved from RNAlater to a saline buffer containing β-mercaptoethanol and homogenized. Crude homogenate was purified by twice centrifuging at 20,000 × g for five minutes. This step was helpful in removing hard material, pile, and fats. In 2014, total RNA was extracted from supernatants using the PureLink RNA Mini Kit (Life Technologies, Carlsbad, California, USA). Samples collected in 2015 were processed using the Maxwell 16 LEV simplyRNA Tissue Kit (Promega, Madison, Wisconsin, USA). Template cDNA was produced by reverse transcription of 1 μg total RNA with an oligo-dT primer (iScript Select cDNA Synthesis Kit, BioRad, Hercules, California, USA).

For each target gene, primers and probes (Table 4) were designed using the Primer3 algorithm52. The mRNA sequences of abaecin, apidaecin, defensin, hymenoptaecin and actin-5C from Bombus impatiens were obtained from GenBank (accession numbers XM_003491496, XM_003491720, XM_003486302, XM_003494885 and XM_003488437, respectively). The specificity of products was verified by dissociation curves as well as sub-cloning and Sanger sequencing. Dual-labeled probes were synthesized by Sigma-Aldrich Custom Products (St. Louis, Missouri).

Table 4: Oligonucleotide primers and dual-labeled probes used to measure AMP expression through multiplex qRT-PCR.

To produce quantitative template standards, plasmid clones of the target sequences were diluted to concentrations of 104, 105, 106, and 107 template molecules per μl53. Salmon sperm DNA (Life Technologies) was added to all standards at 50 ng/μl to approximate the milieu of non-specific sequences in actual cDNA samples. All primer-probe sets performed with high efficiency. Starting numbers of target transcripts were calculated from the mean of technical triplicates, based on the linear regression of standard log10-concentrations. Individual and colony replicates were kept separate for analysis. Data were saved in CSV format and imported for further analysis using R. Data files and the analysis script have been archived with the Dryad Digital Repository (doi:10.5061/dryad.3600k).

Pathogen screening

Infection presents the possibility of a confounding factor in experimental analysis of AMP expression. Therefore, individual abdomens were visually inspected for parasites prior to homogenization. No macroendoparasites or mites were observed. We also screened 18 individuals from the control and high imidacloprid treatments from the day 19 collection, using realtime RT-PCR. The microsporidian Nosema bombi and the trypanosome Crithidia bombi are two of the most common pathogens of B. impatiens. Primers for each parasite were designed to highly expressed house-keeping genes. For N. bombi, methionine aminopeptidase (consensus of GenBank Accessions KF188772–KF188782 and JQ927011) was targeted using the primers Nb’map-161F, CGTCTAAAGAAGCTACGAATGCTG, and Nb’map-257R, TAGCTTCGCATTACTTCGTGGATA. Crithidia bombi was targeting by primers to GAPDH (GenBank Accession GU321192): Cb’gapdh-52F, GCGTACCAGATGAAGTTTGATACG, and Cb’gapdh-147R, AAGCACATCCGGCTTCTTCA. Long primers, spanning these gene sequences, were used to produce cloned gene fragments, for use as standards and positive amplification controls. These primers were parasite-specific and did not amplify from B. impatiens cDNAs known to be parasite-free. Realtime PCR was then used to test B. impatiens cDNAs from our experiment. Strong amplification was observed from positive controls containing parasite sequence. All amplicons were examined for dissociation curves that could confirm the amplification of parasite sequences. None of the 18 samples showed consistent parasite-specific amplification. Therefore, we attribute group differences in AMP expression to imidacloprid exposure, rather than to disease in individual bees.

Statistical analysis

Statistical analyses and their output are presented in their entirety in the Supplementary Methods. Here we provide an overall description of statistical methods. All statistical tests and graphs were rendered using R, version 3.2.254. Group comparisons were made using ANOVA or the Kruskal-Wallis rank sum test after tests of normality and homogeneity of variance. Post-hoc tests for pairwise comparisons were either Tukey’s honest significant difference, following ANOVA, or Dunn’s test with Bonferroni correction. For tests of correlation, we used Spearman’s rank correlation, which does not assume normality and is more robust to small sample sizes. Bonferroni correction was used to adjust p-values when multiple tests were performed on the same data. The log-transcript numbers of each AMP were normalized by subtraction of the value for actin-5c from within the same multiplex reaction. Change in normalized AMP expression over time was calculated by subtracting values for each individual at the final sample time point from the colony mean at the initial time point (Fig. 2E–H).

Additional Information

How to cite this article: Simmons, W. R. and Angelini, D. R. Chronic exposure to a neonicotinoid increases expression of antimicrobial peptide genes in the bumblebee Bombus impatiens. Sci. Rep. 7, 44773; doi: 10.1038/srep44773 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. 1.

    , & Endangered mutualisms: The conservation of plant-pollinator interactions. Annu. Rev. Ecol. Syst. 29, 83–112 (1998).

  2. 2.

    & Threats to an ecosystem service: Pressures on pollinators. Front. Ecol. Environ. 11, 251–259 (2013).

  3. 3.

    , , & Native bees provide insurance against ongoing honey bee losses. Ecol. Lett. 10, 1105–1113 (2007).

  4. 4.

    , , & Decline of bumble bees (Bombus) in the North American Midwest. Biol. Conserv. 142, 75–84 (2009).

  5. 5.

    et al. Patterns of widespread decline in North American bumble bees. Proc. Natl. Acad. Sci. USA 108, 662–667 (2011).

  6. 6.

    et al. Interactions between Nosema microspores and a neonicotinoid weaken honeybees (Apis mellifera). Environ. Microbiol. 12, 774–782 (2010).

  7. 7.

    , , & Bee declines driven by combined stress from parasites, pesticides, and lack of flowers. Science. 347, 1255957 (2015).

  8. 8.

    , & Environment: Overhaul pesticide testing on bees. Nature 497, 188 (2013).

  9. 9.

    , & Sub-lethal effects of imidacloprid on bumblebees, Bombus terrestris (Hymenoptera: Apidae), during a laboratory feeding test. Pest Manag. Sci. 56, 784–788 (2000).

  10. 10.

    et al. A LC/APCI-MS/MS method for analysis of imidacloprid in soils, in plants, and in pollens. Anal. Chem. 75, 2027–2033 (2003).

  11. 11.

    , , & Neonicotinoids in bees: a review on concentrations, side-effects and risk assessment. Ecotoxicology 21, 973–992 (2012).

  12. 12.

    et al. Neonicotinoids, bee disorders and the sustainability of pollinator services. Curr. Opin. Environ. Sustain. 5, 293–305 (2013).

  13. 13.

    et al. Delayed and time-cumulative toxicity of imidacloprid in bees, ants and termites. Sci. Rep. 4, 5566 (2014).

  14. 14.

    The trouble with neonicotinoids. Science 346, 806–807 (2014).

  15. 15.

    et al. Subchronic Exposure of Honeybees To Sublethal Doses of Pesticides: Effects on Behavior. Environ. Toxicol. Chem. 28, 113 (2009).

  16. 16.

    , & The Sublethal Effects of Pesticides on Beneficial Arthropods. Annu. Rev. Entomol. 52, 81–106 (2007).

  17. 17.

    et al. A common pesticide decreases foraging success and survival in honey bees. Science 336, 348–50 (2012).

  18. 18.

    & Insecticide residues in pollen and nectar of a cucurbit crop and their potential exposure to pollinators. Journal of Agricultural and Food Chemistry 60, 4449–4456 (2012).

  19. 19.

    et al. Impact of chronic neonicotinoid exposure on honeybee colony performance and queen supersedure. PLoS One 9, (2014).

  20. 20.

    , & Field realistic doses of pesticide imidacloprid reduce bumblebee pollen foraging efficiency. Ecotoxicology 23, 317–323 (2014).

  21. 21.

    , & Combined pesticide exposure severely affects individual- and colony-level traits in bees. Nature 491, 105–8 (2012).

  22. 22.

    , , & Effects of imidacloprid, a neonicotinoid pesticide, on reproduction in worker bumble bees (Bombus terrestris). Ecotoxicology 21, 1937–1945 (2012).

  23. 23.

    , , & Neonicotinoid Pesticide Reduces Bumble Bee Colony Growth and Queen Production. Science. 336, 351–352 (2012).

  24. 24.

    et al. Seed coating with a neonicotinoid insecticide negatively affects wild bees. Nature 521, 77–80 (2015).

  25. 25.

    et al. Risk assessment for side-effects of neonicotinoids against bumblebees with and without impairing foraging behavior. Ecotoxicology 19, 207–215 (2010).

  26. 26.

    & Chronic impairment of bumblebee natural foraging behaviour induced by sublethal pesticide exposure. Funct. Ecol. 28, 1459–1471 (2014).

  27. 27.

    & Chronic exposure of imidacloprid and clothianidin reduce queen survival, foraging, and nectar storing in colonies of Bombus impatiens. PLoS One 9, e91573 (2014).

  28. 28.

    , , & Pesticide exposure in honey bees results in increased levels of the gut pathogen Nosema. Naturwissenschaften 99, 153–158 (2012).

  29. 29.

    et al. Neonicotinoid clothianidin adversely affects insect immunity and promotes replication of a viral pathogen in honey bees. Proc. Natl. Acad. Sci. 110, 18466–18471 (2013).

  30. 30.

    , , & No effect of low-level chronic neonicotinoid exposure on bumblebee learning and fecundity. PeerJ 4, e1808 (2016).

  31. 31.

    et al. Are bee diseases linked to pesticides? - A brief review. Environment International 89–90, 7–11 (2016).

  32. 32.

    , , , & The neonicotinoids thiacloprid, imidacloprid, and clothianidin affect the immunocompetence of honey bees (Apis mellifera L.). J. Insect Physiol. 86, 40–47 (2016).

  33. 33.

    Antimicrobial peptides of multicellular organisms. Nature 415, 389–395 (2002).

  34. 34.

    , , & Antimicrobial peptides in insects; structure and function. Dev. Comp. Immunol. 23, 329–344 (1999).

  35. 35.

    Cationic peptides: Effectors in innate immunity and novel antimicrobials. Lancet Infect. Dis. 1, 156–164 (2001).

  36. 36.

    , & Dynamics of immune system gene expression upon bacterial challenge and wounding in a social insect (Bombus terrestris). PLoS One 6, e18126 (2011).

  37. 37.

    & Antimicrobial peptides play a functional role in bumblebee anti-trypanosome defense. Dev. Comp. Immunol. 42, 240–243 (2014).

  38. 38.

    , , & Gene expression in honey bee (Apis mellifera) larvae exposed to pesticides and Varroa mites (Varroa destructor). J. Insect Physiol. 58, 1042–1049 (2012).

  39. 39.

    , , , & Differential expression of immune genes of adult honey bee (Apis mellifera) after inoculated by Nosema ceranae. J. Insect Physiol. 58, 1090–1095 (2012).

  40. 40.

    et al. Transcriptome Analyses of the Honeybee Response to Nosema ceranae and Insecticides. PLoS One 9, e91686 (2014).

  41. 41.

    & Bombus impatiens (Hymenoptera: Apidae): An alternative to Apis mellifera (Hymenoptera: Apidae) for lowbush blueberry pollination. J. Econ. Entomol. 94, 609–616 (2001).

  42. 42.

    & Bee (Hymenoptera: Apoidea) diversity and abundance on cranberry in southeastern Massachusetts. Ann. Entomol. Soc. Amer. 88, 334–341 (1995).

  43. 43.

    , , & Impacts of the Use of Nonnative Commercial Bumble Bees for Pollinator Supplementation in Raspberry. J. Econ. Entomol. 104, 107–114 (2011).

  44. 44.

    , & Pollination services provided by bees in pumpkin fields supplemented with either Apis mellifera or Bombus impatiens or not supplemented. PLoS One 8, e69819 (2013).

  45. 45.

    , , & Patterns of range-wide genetic variation in six North American bumble bee (Apidae: Bombus) species. Mol. Ecol. 20, 4870–4888 (2011).

  46. 46.

    , , , & Assessment of chronic sublethal effects of imidacloprid on honey bee colony health. PLoS One 10, e0118748 (2015).

  47. 47.

    & Movement of soil-applied imidacloprid and thiamethoxam into nectar and pollen of squash (Cucurbita pepo). PLoS One 7 (2012).

  48. 48.

    & The host defense of Drosophila melanogaster. Annual Review of Immunology 25 (2007).

  49. 49.

    & Insecticide-induced hormesis and arthropod pest management. Pest Manag. Sci. 70, 690–697 (2014).

  50. 50.

    & Can poisons stimulate bees? Appreciating the potential of hormesis in bee-pesticide research. Pest Manag. Sci. 71, 1368–1370 (2015).

  51. 51.

    et al. Immune pathways and defence mechanisms in honey bees Apis mellifera. Insect Mol. Biol. 15, 645–656 (2006).

  52. 52.

    & Skaletsky. In Bioinformatics Methods and Protocols (eds & ) 132, 365–386 (Humana Press, 2000).

  53. 53.

    In A–Z of Quantitative PCR (ed. ) 87–112 (International University Line Press, 2004).

  54. 54.

    R Core Team. R: A language and environment for statistical computing. (2016).

Download references


The authors wish to thank Serena Graham for assistance with fieldwork. We thank Anne Averill for her helpful comments on an early draft of this manuscript. This work has been supported by research funds from Colby College to D.R.A.

Author information


  1. Department of Biology, Colby College, 5734 Mayflower Hill, Waterville, ME, 04901, USA

    • William R. Simmons
    •  & David R. Angelini
  2. National Human Genome Research Institute, 49 Convent Drive, Bethesda, MD, 20892, USA

    • William R. Simmons


  1. Search for William R. Simmons in:

  2. Search for David R. Angelini in:


Both authors designed the project. D.R.A. supervised and W.R.S. conducted field experiments and gene expression assays. D.R.A. preformed the analyses and prepare figures. Both authors contributed equally to the writing of the main text.

Competing interests

The authors declare no competing financial interests.

Corresponding author

Correspondence to David R. Angelini.

Supplementary information

About this article

Publication history






By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.