
==== Front
R Soc Open Sci
R Soc Open Sci
RSOS
royopensci
Royal Society Open Science
2054-5703
The Royal Society

rsos231654
10.1098/rsos.231654
10021002100137515Chemistry
Research Articles
Phytochemical profiles of honey bees (Apis mellifera) and their larvae differ from the composition of their pollen diet
Phytochemical profiles of honey bees (Apis mellifera) and their larvae differ from the composition of their pollen diet
https://orcid.org/0000-0001-6383-7674
Vidkjær Nanna Hjort 1 2 Conceptualization Data curation Formal analysis Funding acquisition Investigation Methodology Project administration Resources Validation Visualization Writing – original draft Writing – review and editing nanna.vidkjaer@gmail.com

Laursen Bente B. 2 Investigation Validation benteb.laursen@agro.au.dk

Kryger Per 2 Conceptualization Investigation Methodology Resources Writing – review and editing per.kryger@agro.au.dk

1 Department of Biology, University of Copenhagen , Copenhagen, Denmark
2 Department of Agroecology, Aarhus University , Slagelse, Denmark
Electronic supplementary material is available online https://doi.org/10.6084/m9.figshare.c.7454632.

This article has been edited by the Royal Society of Chemistry, including the commissioning, peer review process and editorial aspects up to the point of acceptance.

9 2024
25 9 2024 September 25, 2024
25 9 2024 September 25, 2024
11 9 23165407 11 2023 November 7, 2023
19 3 2024 March 19, 2024
13 8 2024 August 13, 2024
© 2024 The Authors.
2024
https://creativecommons.org/licenses/by/4.0/ Published by the Royal Society under the terms of the Creative Commons Attribution License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original author and source are credited.

Pollen and nectar consumed by honey bees contain plant secondary metabolites (PSMs) with vital roles in plant–insect interactions. While PSMs can be toxic to bees, they can also be health-promoting, e.g. by improving pesticide and pathogen tolerances. As xenobiotics, PSMs undergo post-ingestion chemical modifications that can affect their bioactivity and transmission to the brood. Despite the importance of understanding honey bee PSM metabolism and distribution for elucidating bioactivity mechanisms, these aspects remain largely unexplored. In this study, we used HPLC-MS/MS to profile 47 pollen PSMs in honey bees and larvae. Both adult bees and larvae had distinct PSM profiles that differed from their diet. This is likely due to post-ingestion metabolism and compound-dependent variations in PSM transmission to the brood via nurse bee jelly. Phenolic acids and flavonoid aglycones were most abundant in bees and larvae, whereas alkaloids, cyanogenic glycosides and diterpenoids had the lowest abundance despite being consumed in higher concentrations. This study documents larval exposure to a variety of PSMs for the first time, with concentrations increasing from early to late larval instars. Our findings provide novel insights into the post-ingestion fate of PSMs in honey bees, providing a foundation for further exploration of biotransformation pathways and PSM effects on honey bee health.

bee health
; flavonoids
; alkaloids
; phenolic acids
; post-ingestion metabolism
; plant secondary metabolites
Danmarks Frie Forskningsfond http://dx.doi.org/10.13039/501100004836
==== Body
pmc1. Introduction

Plants produce an extraordinary array of secondary metabolites (PSMs, also known as plant specialized metabolites) with important roles in defences against herbivores and pathogens, as well as in plant signalling [1,2]. These phytochemicals are also present in the pollen and nectar consumed by bees [3,4], but their role in plant–pollinator interactions is less obvious. PSMs exhibit a wide range of bioactivities, including antiviral and antimicrobial properties. While some PSMs can be toxic [5,6], co-evolutionary mutualism seems to enable insect pollinators to derive benefits from these compounds, particularly in their defences against pathogens [7–10]. Recent pollinator declines have led to an increasing number of studies on how PSMs influence bee health. Although only a few PSMs have been tested, these studies suggest a multifaceted role including toxic, antiviral [11,12] and immunostimulatory activity [13–15], improved pathogen defences [16–21], increased longevity [17,18] and changes in gut microbial composition and abundance [22]. Some flavonoids and phenolic acids enhance honey bees’ pesticide tolerance and detoxification by inducing the expression of genes regulating the cytochrome P450 detoxification system [10,15,23–26]. PSMs have primarily been tested on adult bees, leaving health effects and toxicity levels in the brood underexplored. For species like honey bees, which feed their larvae glandular secretions rather than pollen and nectar directly, brood PSM exposure remains largely unexplored as it depends on the extent to which pollen and nectar PSMs are transferred to these secretions.

PSMs are xenobiotics to bees and metabolized after ingestion by the bees’ own enzymatic detoxification systems [27,28] with contributions from gut microorganisms [29,30]. Despite the increasing amounts of research showing diverse PSM functions in bee health, post-ingestion chemical reactions are poorly understood in bees as well as in other insects. Metabolization pathways, detailing the transformation steps of PSMs, likely differ based on molecular structure and properties similar to mammalian PSM metabolism [31] but are known for only a few PSMs in a small number of species. Knowing how various PSMs are taken up, metabolized and distributed within bees is important for understanding and predicting PSM bioactivities and unravelling mechanisms underlying specific health effects. This was for instance recently exemplified in bumble bees for a PSM with antiparasitic activity where the active compound was a metabolite formed in the bees while the ingested parent PSM was inactive [32]. This also emphasizes that laboratory assays testing the direct activity of PSMs against cultured pathogens (e.g. [32]) may not accurately reflect the in vivo effects, as these assays do not account for post-ingestion metabolization or gut PSM uptake and distribution within bees.

For honey bees, which feed their larvae jelly produced in the specialized hypopharygeal glands of nurse bees [33,34], the post-ingestion fate of PSMs can also influence transmission to the brood. Early larval instars receive only jelly, while older larvae destined to become worker bees also receive small amounts of pollen and honey [33,34]. Nurse bees must consume large quantities of bee bread [33]—a mix of pollen, honey, nectar and glandular secretions [35]—to produce jelly. This creates a potential, but underexplored, pathway for PSM transmission to the brood via jelly. The exact mechanism of PSM transfer from diet to jelly remains unknown but the PSM profile of jelly likely differs from that of the diet. Some PSMs may undergo metabolization before being passed to the glands, while others may not be transmitted at all. To date, only a few studies document PSMs in jelly [36,37] while none have tested the PSM content in honey bee brood. If the brood is exposed to a wide range of PSMs, understanding how PSMs influence larval health is essential—not only for assessing impacts on the individual larva but also for understanding the effects on colony health and rejuvenation.

In this study, we exposed miniature honey bee colonies to a field-realistic clover pollen diet containing flavonoids and phenolic acids, which are ubiquitous in pollen and nectar. To introduce a broader range of dietary PSM compound classes, we fortified this diet with alkaloids, diterpenoids and cyanogenic glycosides produced by fewer plant species. Our objectives were to (i) identify dietary PSMs present in adult honey bees and larvae, (ii) determine how the PSM composition changes across larval developmental stages, and (iii) compare PSM profiles between adult bees and larvae. We hypothesized that the PSM profiles of both adult bees and larvae would differ qualitatively from the diet and vary across different PSM molecular structures due to metabolization differences. Our aims were to determine the PSM content of larvae for the first time and provide new insights into the transmission of PSMs from diet to adult bees and brood, including comparing the transmission of PSMs that are omnipresent in pollen and nectar with compounds produced by fewer plant species. The results provide new insights to guide detailed mapping of key PSM metabolization pathways and PSM impact on larval health.

2. Material and methods

2.1. Chemicals

Details about the chemicals used for extractions and as analytical standards for the HPLC-MS/MS analyses are included in the electronic supplementary material (table S1).

2.2. Honey bees

Six small Apis mellifera L. colonies (~750–1100 bees per colony) were kept in miniature hives (24 × 15 × 15 cm) (Mini-mating nucs; Apidea, Switzerland) consisting of a queen, worker bees and brood [38,39]. The hives originated from sister queens hatched in June and mated in early July, establishing small colonies for this experiment in September. The six colonies were selected based on size similarity and comparable amounts of brood. Three hives were used for the feeding experiment focusing on PSMs in larval stages (§2.4), while the other three provided newly hatched bees for a separate experiment measuring dietary PSMs in adult bees (§2.5). All miniature hives were kept in the apiary of the Department of Agroecology, Aarhus University, Flakkebjerg, Denmark.

2.3. Pollen diets

The diet consisted of white clover (Trifolium repens) pollen fortified with a series of PSMs. Clover pollen PSMs have not previously been analysed, but existing knowledge on PSMs in white clover leaves, stems, flowers and roots suggests that the pollen contains a range of flavonoids and phenolic acids [40–45] (see §2.6.1). To expand the range of compound classes, we fortified the clover pollen with six additional PSMs: gelsemine (indole alkaloid), atropine (tropane alkaloid), senkirkine (pyrrolizidine alkaloid), amygdalin (cyanogenic glycoside), triptolide (diterpenoid epoxide) and methyllycaconitine (diterpenoid alkaloid) (systematic names are included in electronic supplementary material, table S2). These compounds were selected based on three criteria: (i) they represent PSM compound classes not naturally present in white clover, (ii) their natural concentrations in pollen and/or nectar are reported in the literature (see [46–53]), and (iii) they are commercially available as isolation from natural sources in the needed amounts was not feasible. Toxic effects have been reported for several of these PSMs [5,51,54–56], with details on toxicity levels and bioactivities in adult bees summarized in Vidkjær et al. [57]. No studies have tested these compounds in bee larvae, and larval toxicity levels, health effects and capacity for metabolizing these compounds are unknown. The concentrations in the fortified diet were within the natural range for pollen and/or nectar and below toxicity thresholds for adult honey bees (summarized in Vidkjær et al. [57]).

Corbicular white clover pollen was collected with traps mounted at the entrances of honey bee hives located in clover fields at the Department of Agroecology, Aarhus University, Flakkebjerg, Denmark, and kept at −20°C. Palynological analyses confirmed that the pollen was 92% Trifolium repens. To prepare experimental diets, 15 g of corbicular pollen was mixed with 1 ml Apiinvert® (28% water and 72% sugar: 30% saccharose, 31% dextrose, 39% fructose) and 0.4 ml water using a pestle and mortar. The fortified pollen diet was then prepared by adding 1.5 mg gelsemine, 0.3 mg atropine, 0.75 mg amygdalin, 0.045 mg senkirkine, 0.015 mg methyllycaconitine and 0.015 mg triptolide. These compounds were first dissolved in either methanol, ethanol or ethyl acetate to create stock solutions (300 µg ml−1 to 22.5 mg ml−1) kept at −20°C. When the diet was prepared, the individual PSMs were mixed by adding 50–82 µl of each stock solution (463 µl in total) to 0.5 ml ethanol. This mixture was then added to the clover pollen paste and thoroughly mixed with a pestle and mortar to ensure uniform distribution. Approximately 5 g of the resulting pollen paste was then transferred to sterile plastic well plates and left at room temperature for 1 h to allow solvent evaporation. The well plates were then attached to new, unused wooden frames without wax combs with elastic bands and mounted in the miniature hives (electronic supplementary material, figure S1). The pollen diets were prepared and replaced every 24 h to minimize compound degradation and monitor food intake.

2.4. Feeding experiments with small honey bee hives to measure plant secondary metabolites in larval stages

Three miniature hives (see §2.2), H1, H2 and H3, with adult bees, brood and a queen, were used for this experiment. One week before the experiment started, the nectar frames were removed and replaced with plastic well-plates containing PSM-free Apifonda® (85% sugar, mainly saccharose; approx. 300 g per hive). Three days before the experiment, hive entrances were closed, and pollen frames were removed to prevent unintended PSM exposure. The experiment began with feeding the sealed hives a pure clover pollen diet for 2 days. On day three, feeding with the fortified clover pollen diet started and continued for 3 days. The experiment ended on day six. By this time, the hives had been sealed for 8 days, the average duration for a honey bee egg to hatch and reach the pupae stage.

To collect the larvae, the hives were anaesthetized with CO2, and the number of bees counted (H1: 1120 bees; H2: 901 bees; H3: 747 bees). Larvae in the third, fourth and fifth instar (separated based on size) were gently collected, washed with Milli-Q water to remove PSM-containing jelly from the cuticle, and sacrificed by freezing. The numbers of larvae collected from each nucleus hive were H1: third instar, 85; fourth instar, 33; fifth instar, 35; H2: third instar, 51; fourth instar, 42; fifth instar, 19; H3: third instar, 37; fourth instar, 26; fifth instar, 17. Given the larvae’s age distribution, they must have hatched after the hives were sealed within the timeframe of the experiment. Five fifth instar, five fourth instar and ten third instar larvae were kept for analyses of pollen grains in their gut. For chemical analyses, three sub-samples of each larval instar were prepared for each hive by pooling the following numbers of larvae: third instar, n = 9; fourth instar, n = 7; and fifth instar, n = 4. This resulted in nine larval samples per hive and a total of 27 samples for chemical analyses. These samples were freeze-dried for 72 h (Heto Drywinner, Birkerød, Denmark). Metal beads were then added to the dried larval samples, which were pulverized by vibration for 30 s at 1500 r.p.m. using a Geno/Grinder (SPEX Sample Prep 2010, Metuchen, NJ, USA).

2.5. Feeding experiment with adult honey bees

PSM profiles of adult honey bees were assessed in a separate setup. Flavonoids and phenolic acids are omnipresent in pollen and nectar [26,58], and our previous analyses (unpublished data) show that adult honey bees retain these compounds even after consuming a pure PSM-free sucrose diet for several days. In the experimental miniature hives (§2.4), we could not distinguish young adult bees hatched after the hives were sealed from older foragers pre-exposed to pollen and nectar that were in the hives before the beginning of the experiment. To minimize background levels of flavonoids and phenolic acids and to analyse PSMs in young bees that have the highest pollen consumption during their maturation into nurses, we performed this experiment in a separate setup.

Newly hatched adult bees for this experiment came from the three miniature colonies H4, H5 and H6 (see §2.2). Before starting the feeding experiment, we sealed the hives and replaced pollen and nectar frames with PSM-free Apifonda® (85% sugar, mainly saccharose; approx. 300 g per hive) and pure clover pollen diet (see §2.3). After two weeks, frames with capped brood cells were moved to a dark incubator at 35°C and 30–50% humidity. Newly emerged bees were collected after 2 days and combined into one queenless miniature hive (24 × 15 × 15 cm) (Mini-mating nuc; Apidea, Switzerland) containing bees from H4, H5 and H6. A queen pheromone strip was used to simulate the presence of a queen. To replicate PSM exposure from the larval experiment (§2.4), the young adult bees were fed pure clover pollen diet for 2 days, followed by fortified clover pollen diet for 3 days, with additional access to Apifonda®. Diets were replaced daily to prevent PSM degradation and measure pollen intake. The mean daily intake was 5.6 ± 2.9 mg per bee. On day six, feeding was terminated, and the bees were anaesthetized with CO2, counted (285 bees) and sacrificed by freezing. Bees were washed with Milli-Q water to remove pollen diet residues from the cuticle, then divided into five sub-samples of 50 bees each. These samples were freeze-dried for 72 h (Heto Drywinner, Birkerød, Denmark), then metal beads were added, and the samples were pulverized by vibration for 30 s at 1500 r.p.m. using a Geno/Grinder (SPEX Sample Prep 2010, Metuchen, NJ, USA).

2.6. Sample extraction

2.6.1. Pollen

The PSM composition of white clover pollen has not previously been assessed. The 29 flavonoids and 12 phenolic acids targeted by our HPLC-MS/MS methods (see electronic supplementary material, table S2, for a complete list including systematic names) were selected based on existing knowledge of PSMs in white clover plants [42,45,59].

Flavonoids and phenolic acids were extracted from 100 mg clover pollen with 4 ml of 7:3 methanol:water by ultrasonication for 1 h at 30°C. The extract was centrifuged (12 min, 4°C, 4500 r.p.m.), the supernatant was removed and the extraction was repeated. The two extracts were combined and stored at −20°C. The extraction protocol was based on quantitative analyses of flavonoids and phenolic acids in pollen [60] and bee bread [61] established by others.

The six additional PSMs artificially added to the pollen diet were not quantified as they were added to the diet in specific known concentrations. These concentrations were gelsemine (100 ng mg–1), atropine (20 ng mg–1), amygdaline (50 ng mg–1), senkirkine (3 ng mg–1), methyllycaconitine (1 ng mg–1) and triptolide (1 ng mg–1).

2.6.2. Adult honey bees

Extractions of gelsemine, atropine, amygdaline, senkirkine, methyllycaconitine and triptolide from the five pools of adult honey bees followed previously published and validated protocols using 35 mg of the dried bee samples [57].

Flavonoids and phenolic acids were extracted from 35 mg dried bee samples with 1 ml of 7:3 methanol:water by shaking for 1 h using an Intelli-Mixer (60 r.p.m.). The extracts were centrifuged for 12 min (4°C, 4500 r.p.m.) and stored at −20°C. Details on method development and performance are included in §2.8.

2.6.3. Larvae

Extractions of the pooled larval samples were performed with 10 mg of the third instar and 15 mg of the fourth and fifth instar samples. The extraction protocols for gelsemine, atropine, amygdaline, senkirkine, methyllycaconitine and triptolide were optimized from protocols developed for adult honey bees [57]. Initial tests with the adult bee protocols did not provide adequate results for larval samples and were modified as described in the following text. Details on method adaptation to larval samples and performance evaluation are included in §2.8.

Amygdalin and atropine were extracted with 1 ml of 7:3 methanol:water + 0.2% acetic acid, while gelsemine, senkirkine and methyllycaconitine were extracted with 1 ml of 1:1 methanol:water. The shaking extractions were performed using an Intelli-Mixer (60 r.p.m.) for 1 h. The extracts were centrifuged for 12 min (4°C, 4500 r.p.m.) and stored at −20°C. Flavonoids and phenolic acids were extracted by 1 h shaking extraction with 1 ml 7:3 methanol:water as described in §2.6.2 for adult honey bee samples.

2.7. HPLC-MS/MS analyses and quantification

Prior to HPLC-MS/MS analyses, all larvae, adult bee and pollen extracts were diluted with water and filtered using a syringe filter (Kinesis KX PTFE syringe filter 13 mm, 0.22 μm, Mikrolab, Aarhus, Denmark).

The 12 phenolic acids and the six artificially added PSMs (electronic supplementary material, table S2) were quantified using an Agilent 1260 Infinity HPLC (Agilent Technologies, Glostrup, Denmark) coupled with a Sciex (Copenhagen, Denmark) 4500 QTRAP mass spectrometer equipped with electrospray ionization and operated in multiple reaction monitoring (MRM) mode. The instrument was controlled by Analyst software version 1.6.3. Nitrogen was used as source and collision gas. The compound-dependent mass spectrometer parameters (declustering potential (DP), entry potential (EP), cell entry potential (CEP), collision energy (CE) and cell exit potential (CXP)) that were optimized by infusing the pure compound directly into the mass spectrometer source are listed in electronic supplementary material, table S3, together with the two MRM transitions monitored for each compound. The MRM giving the highest response was used for quantification, and the other was used as a qualifier to ensure correct identification. The phenolic acids were analysed in negative mode with chromatographic separation on a Phenomenex Kinetex Polar C18 column (100 × 2.1 mm, 2.6 μm, 100 Å; Phenomenex, Værløse, Denmark) with a flow rate of 0.4 ml min–1 and mobile phases A: 92% water + 4% methanol + 4% 2-propanol + 20 mM acetic acid and B: 92% acetonitrile + 4% methanol + 4% 2-propanol + 20 mM acetic acid. The gradient was as follows: 0–2 min: 100% A; 2–3.6 min: 100–54.4% A; 3.6–3.8 min: 54.4–16.3% A; 3.8–4.5 min: 16.3% A; 4.5–4.7 min: 16.3–100% A; 4.7–8 min: 100% A. The column oven was set to 30°C, and the injection volume was 10 µl. The mass spectrometer source parameters were curtain gas (CUR), 35 psi; collision gas (CAD), medium; temperature (TEM), 600°C; ion source gas 1 (GS1), 90 psi; ion source gas 2 (GS2), 90 psi; and ionspray voltage (IS), −4300 V. Standard curves were recorded in the concentration interval 0.048–800 ng ml–1. Gelsemine, atropine, senkirkine, methyllycaconitine and triptolide were analysed in positive mode and amygdalin in negative mode using chromatographic conditions and mass spectrometer source parameters as described in Vidkjær et al. [57]. Standard curves were recorded in the concentration interval 0.0015–50 ng ml–1 for amygdalin, gelsemine, atropine, senkirkine and methyllycaconitine and 0.006–2.5 ng ml–1 for triptolide. Quantifications were performed in Analyst software version 1.6.3.

The 29 flavonoids (electronic supplementary material, table S2) were quantified using an Agilent 1260 Infinity HPLC (Agilent Technologies, Glostrup, Denmark) coupled with a Sciex (Copenhagen, Denmark) 3200 QTRAP mass spectrometer equipped with electrospray ionization and operated in MRM mode. The instrument was controlled by Analyst software version 1.6.3. The chromatographic conditions and mass spectrometer source parameters were as previously published [62]. The optimized compound-dependent mass spectrometer parameters (DP, EP, CE, CXP) are listed in the electronic supplementary material, table S3, together with the two MRM transitions monitored and the ionization mode applied for each flavonoid. Standard curves were recorded in the concentration interval 0.39–800 ng ml–1, and quantifications were performed in Analyst software version 1.6.3.

2.8. Analytical method development and validation

The analytical protocols for adult honey bees and their larvae were developed by spiking known concentrations of pure PSMs into dried pulverized honey bee and larval samples to test different extraction protocols. The bees and larvae used for method development did not originate from the experimental hives and were confirmed to be free of the targeted PSMs by HPLC-MS/MS. All six artificially added PSMs and a subset of flavonoids and phenolic acids were used to develop and evaluate the performance of the analytical protocols.

To validate the optimized analytical protocols, accuracy and precision were tested by spiking two separate concentrations of the pure PSMs into dried pulverized adult honey bees (n = 5) and larvae (n = 6) free from the targeted PSMs. Recovery percentages, representing the percentage of the theoretical concentration measured in the spiked samples, were calculated to evaluate method accuracy, while precision was assessed by the relative standard deviation in per cent (RSD%) between replicate samples. The limits of detection (LOD) and quantification (LOQ) were set as the standard concentrations for which the signal-to-noise ratios (S/N) were equal to or above 3 and 10, respectively. Method validation details and data are provided in the electronic supplementary material (tables S4 and S5). Briefly, the recovery percentages for flavonoids and phenolic acids ranged from 51 to 98% for adult bees and 59–116% for larvae. RSD% were 4–16% and 3–17% for adult bees and larvae, respectively. The methods for the six PSMs artificially added to the diet were previously validated for adult honey bees, as detailed in Vidkjær et al. [57]. The methods adopted for analysing these compounds in larval samples were validated as part of this experiment, with recovery percentages ranging from 63 to 134% and RSD% from 2 to 12%. The recovery percentages were not used to correct the concentrations measured in adult bees and larvae.

2.9. Microscopic analyses of pollen in larvae guts

Fourth and fifth instar larvae were homogenized individually (n = 5) in an Eppendorf® tube with a fitting micropestle® and 100 µl of Milli-Q water. Due to their smaller size, the third instar larvae were homogenized in a pool of ten larvae with 200 µl of Milli-Q water. Microscopic examinations of the homogenates were done with phase contrast at 100× to determine the presence or absence of pollen grains (electronic supplementary material, figures S2–S4). Identification of the pollen species as clover was done at 300× magnification (electronic supplementary material, figure S5).

2.10. Data analyses and statistics

Data analyses and statistics were performed using R version 4.0.4 and the R Base Package. Changes in PSM diversity across diet, adult bees, and larvae were assessed by comparing the number of PSMs detected in these samples by one-way analyses of variance (ANOVA; α = 0.05) followed by Tukey’s test for post hoc analyses (figure 2; see electronic supplementary material, table S6, for statistical output). FDR-corrected p-values were used to adjust for multiple comparisons. Principal component analyses (PCA; data centred and scaled) were performed to explore changes in PSM composition between the diet and the adult bees/larvae as well as between adult bee and larval stages. Prior to PCA, concentration data were converted to percentages of the total PSM concentrations in each sample (electronic supplementary material, table S7). This approach allowed for the comparison of the compositional distribution of individual PSMs across pollen, adult bees and larvae rather than comparing raw concentrations, which varied significantly between the sample groups. The linear relationship between dietary PSM concentrations and the concentrations measured in the larvae and adult bees was explored for each of the three PSM groups (flavonoids, phenolic acids, added PSMs) by plotting the diet concentrations versus the concentrations in adult bees and larvae (third, fourth and fifth instar larvae plotted separately) (electronic supplementary material, figure S6). Linear regression was used to fit the data points, and linearity was assessed using the r 2 value. To test how the concentrations of individual PSMs change with larval age, the fold-change (FC) values between instars (third versus fourth, fourth versus fifth and third versus fifth) were calculated for each PSM (electronic supplementary material, table S8). Only larvae from the same miniature hive were compared, as the specific FC values differed between the hives. For each hive, one-way ANOVAs followed by Tukey’s test for post hoc analyses (α = 0.05) were used to determine if the concentration changes represented by the FC values were significant (electronic supplementary material, table S9). Statistical outputs are included in electronic supplementary material, table S10. FDR-corrected p-values were used to adjust for multiple comparisons.

3. Results and discussion

3.1. Flavonoids and phenolic acids in white clover pollen

Approximately 90% of the pollen flavonoids were O-glycosides of luteolin (luteolin−3′,7-di-O-glucoside, hereafter luteolin-di-glc), quercetin (isoquercetrin and hyperoside) and kaempferol (astragalin) (figure 1). Rutin, quercetin, naringenin, nicotiflorin, quercetin-xyl-gal (quercetin−3-O-{β-d-xylosyl-(1→2)-β-d-galactoside}), kaempferol-rha-glc (kaempferol−7-O-{α-l-rhamnosyl-d-glucoside}), medicarpin, and genistin had low concentrations from 0.06 to 2.7 ng mg–1 pollen (0.16–7.2% of the pollen flavonoids). Biochanin A, formononetin, luteolin, kaempferol-rha-xyl-gal (kaempferol−3-O-{β-d-rhamnosyl-(1→6)-[β-d-xylosyl-(1→2)]-β-d-galactoside}), quercetin-rha-xyl-gal (quercetin−3-O-{β-d-rhamnosyl-(1→6)-[β-d-xylosyl-(1→2)]-β-d-galactoside}) and isovitexin were present at trace levels (figure 1; electronic supplementary material, table S11). Quercitrin, daidzein, saponarin, sissotrin, luteolin-glc (luteolin−4′-O-glucoside), puerarin, daidzin, coumestrol, genistein and kaempferol, which have previously been identified in flowers, leaves or stems of white clover plants [42], were not detected in the pollen. The pollen phenolic acids were primarily composed of three cinnamic acid derivatives, p-coumaric, ferulic and caffeic acids, and two benzoic acid derivatives, 4-hydroxybenzoic and salicylic acids (figure 1) (collectively ~88%). Salicylic, vanillic, gallic, protocatechuic, cinnamic and sinapic acids had low concentrations from 0.1 to 0.99 ng mg–1 pollen (0.61–6.1% of the phenolic acids) (figure 1; electronic supplementary material, table S11) while chlorogenic and syringic acids were detected at trace levels (electronic supplementary material, table S11).

Figure 1. Plant secondary metabolites (PSMs) in pollen diet, adult honey bees and larvae. Heatmap showing the distribution of PSMs in pollen diet, adult bees and larvae from the three miniature hives (H1, H2, H3). The colours indicate the per cent distribution of each compound, calculated as the mean percentage of the total concentration of PSMs within the three compound groups: flavonoids, phenolic acids and PSMs artificially added to the clover pollen diet. The numbers represent the summed mean concentrations of each compound group within the samples. Only the most abundant flavonoids and phenolic acids, with concentrations ≥5% in at least one sample, are included. Data for the remaining flavonoids and phenolic acids are provided in electronic supplementary material, tables S11 and S12, along with individual concentrations of all PSMs analysed by HPLC-MS/MS.

PSMs in pollen diet, adult bees, and larvae.

3.2. Plant secondary metabolites from pollen recovered in adult honey bees

The diversity of PSMs in the young adult honey bees was significantly higher (figure 2a ) than the pollen diet, but the difference was driven by two phenolic acids with low bee abundance (≤5%) that were only detected at trace levels in the pollen (chlorogenic acid and syringic acid). When PSMs with ≤5% abundance were removed, the difference in PSM diversity became non-significant (figure 2b ). The PCA showed that the PSM compositions of adult bees differed from the pollen diet (figure 3). Flavonoid glycosides and the six artificially added PSMs were dominant in the pollen, while phenolic acids and flavonoid aglycones had the highest concentrations in the bees (figure 3). Approximately 90% of the bee flavonoids were the aglycones, quercetin and kaempferol. Quercetin was only a minor component of the pollen diet (2%), whereas kaempferol was not detected (figure 1 and electronic supplementary material, table S11). We conclude that the bee flavonoid aglycones originate from dietary O-glycosides of quercetin and kaempferol that were de-glycosylated in the bees following ingestion. The main pollen flavonoid, luteolin-di-glc, was not recovered in the bees, and the corresponding aglycone, luteolin, was only a minor component (~1%). The remaining pollen flavonoids were either absent or had low concentrations in bees (0.2–3.5%) (figure 1 and electronic supplementary material, table S11).

Figure 2. Plant secondary metabolite (PSM) diversity differs between pollen diet, adult honey bees and larvae. Diversity is measured as the number of detected PSMs: (a) includes all PSMs and (b) compares the most abundant PSMs with concentrations ≥5% in any of the sample groups. Bars represent mean values, with error bars indicating standard deviation. Compound diversity was evaluated separately for each of the four PSM groups—all PSMs, added PSMs, flavonoids and phenolic acids—by one-way ANOVAs (electronic supplementary material, table S6). Letters above the bars denote significance and for each PSM group, bars having letters in common are not significantly different.

PSM diversity in pollen diet, adult bees, and larvae.

Figure 3. Comparison of plant secondary metabolite (PSM) profiles in pollen diet, adult honey bees and larvae. (a) The PCA scores plot shows distinct PSM profiles of adult bees and larvae that differ from their pollen diet (adult bees, n = 5 pools; larvae, n = 3 pools per hive and instar; pollen, n = 3). The PCA loadings plot (b) shows that the pollen diet is characterized by flavonoid glycosides and the six artificially added PSMs, while adult bees and larvae primarily contain phenolic acids and a few flavonoid aglycones. The numbers correspond to individual PSMs: 1, astragalin; 2, genistin; 3, hyperoside; 4, isoquercitin; 5, kaempferol; 6, kaempferol-rha-glc; 7, luteolin; 8, luteolin-di-glc; 9, medicarpin; 10, naringenin; 11, nicotiflorin; 12, pratensein; 13, quercetin; 14, quercetin-xyl-gal; 15, rutin; 16, 4-hydroxybenzoic acid; 17, caffeic acid; 18, chlorogenic acid; 19, cinnamic acid; 20, coumaric acid; 21, ferulic acid; 22, gallic acid; 23, protocatechuic acid; 24, salicylic acid; 25, sinapic acid; 26, syringic acid; 27, vanillic acid; 28, amygdalin; 29, atropine; 30, gelsemine; 31, methyllycaconitine; 32, senkirkine; 33, triptolide.

Comparison of PSM profiles in pollen diet, adult bees and larvae.

All pollen phenolic acids were recovered in adult bees, but the compound per cent distribution differed from the diet (figures 1 and 3). The bees mainly contained protocatechuic, vanillic and caffeic acids (~87% of the total phenolic acid concentration) (figure 1). Of these, only caffeic acid had high pollen abundance (16%), whereas protocatechuic and vanillic acids were minor components (≤6%). The remaining nine pollen phenolic acids had low bee concentrations (0.05–3.9%), including the pollen phenolic acid with the highest concentration, p-coumaric acid (figure 1; electronic supplementary material, table S11). All six PSMs artificially added to the diet were recovered in the adult bees (figure 1 and electronic supplementary material, table S11). Gelsemine, amygdalin and atropine that were added to the pollen diet in higher concentrations were more abundant in the bees than the PSMs added in lower concentrations (senkirkine, methyllycaconitine and triptolide) (figure 1). However, when examining the relationship between diet and bee concentrations, the trend was not linear. The r 2 value was 0.76 for the added PSMs (electronic supplementary material, figure S6) but atropine was for instance detected in the bees in higher concentrations than amygdalin even though the diet atropine concentration was 2.5 times lower than amygdalin. Furthermore, the content of methyllycaconitine in bees was approximately six times higher than senkirkine and triptolide, although the diet methyllycaconitine concentration was the same as triptolide and three times lower than senkirkine (figure 1 and electronic supplementary material, table S11). For the clover flavonoids and phenolic acids, there were no indications of a linear relationship between diet and bee concentrations, and the r 2 values were 0.14 and 0.10, respectively (electronic supplementary material, figure S6). Variations in uptake and/or differences in post-ingestion metabolization between individual PSMs likely explain why high dietary concentrations of specific PSMs do not necessarily result in high bee concentrations. These results are in agreement with our previous findings [57].

Flavonoids and phenolic acids have been recovered in several insect species. Flavonoids are predominantly recovered as glycosides [63–67], which may originate from the diet or be formed post-ingestion by glycosidation. The conjugation of dietary flavonoid aglycones can function as an insect detoxification mechanism [68]. Contrary to these observations, our data show that honey bees extensively de-glycosylate flavonoids post-ingestion (figure 4). Honey bees secrete glycoside hydroxylases from their hypopharynx glands into their mouth. These enzymes are transmitted to the midgut, where they may catalyse de-glycosylation of flavonoids [74,75]. Gut microorganisms may further contribute to PSM metabolization and work in conjunction with the honey bees’ own enzyme systems [29,30]. The grasshopper, Tropidacris collaris, also de-glycosylates dietary flavonoid conjugates and this was suggested to serve as an important carbohydrate source [76]. It is not known if flavonoid de-glycosylation has nutritional value for honey bees, and nectar serves as their primary carbohydrate source. However, de-glycosylation of dietary flavonoids may provide additional carbohydrates not found in nectar, which predominantly consists of sucrose, fructose and glucose [77]. Some flavonoids are known to positively impact honey bee disease resistance and pesticide tolerances and understanding their metabolic fate can facilitate mechanistic insight. We for example recently linked several dietary flavonoid glycosides to enhanced pesticide detoxification in honey bees [26]. Based on the results of the present experiment, we hypothesize that this increase in pesticide metabolism was caused by the respective aglycones formed post-ingestion that increase the expression of P450 genes similarly to quercetin [23].

Figure 4. Proposed metabolization pathways for the most abundant clover pollen phenolic acids and flavonoids in adult honey bees. Unbroken arrows represent well-established biotransformation pathways, while dashed arrows indicate hypothesized pathways. The proposed pathways are based on mammalian plant secondary metabolite (PSM) metabolism [69–73], as detailed knowledge of PSM biotransformations in bees and other insects is currently lacking.

Proposed metabolization pathways for the most abundant clover pollen phenolic acids and flavonoids in adult honey bees.

The main pollen flavonoid, luteolin-di-glc, and its aglycone, luteolin, were only detected in low concentrations in the bees (figure 1). We hypothesize that the low luteolin abundance, compared with quercetin and kaempferol, arises from differences in how honey bees metabolize the structurally different flavonol (quercetin and kaempferol) and flavone (luteolin) aglycones (figure 4). Flavonones may be converted to downstream metabolites faster than the two flavonols, leading to lower luteolin abundance. This is in agreement with our previous documentation of substantial variation in the degradation ratios of structurally different PSMs in honey bees [57]. The metabolic fate of the flavonoid ring system remains virtually unexplored in insects. Elucidation of full pathways would require a different analytical approach from that utilized in this experiment. However, mammalian degradation of the flavonoid aglycone ring system produces various phenolic acids, including protocatechuic, caffeic and vanillic acids [69–72] that were targeted by our analyses. Protocatechuic acid, a well-established mammalian metabolite of luteolin [69–71], was the most abundant phenolic acid in the bees (47%) but only a minor component of the diet (~1%). We hypothesize that the high concentrations we documented in the adult honey bees originate from ring-system cleavage of dietary flavonoids in the bees, mainly from luteolin-di-glc (figure 4). The other luteolin metabolites, caffeic and vanillic acids, were the second and third most abundant phenolic acids in the bees, but these were ingested in higher concentrations. Metabolization of dietary cinnamic acids may also have contributed to the formation of protocatechuic and vanillic acids in the bees as these compounds are also known metabolites of caffeic and ferulic acids, respectively (figure 4) [72,73]. Honey bees possess flavonoid-metabolizing cytochrome P450 enzymes [78] but metabolites originating from flavonoid ring-system cleavage were recently identified in the honey bee gut [29], suggesting that gut microorganisms may work in synergy with the bees’ own enzyme system. Based on our data, it is not possible to determine where in the bees these metabolites were formed.

Metabolization pathways for the PSMs artificially added to the clover pollen diet have not yet been clarified in honey bees or any other insect except for amygdalin metabolism, which was recently determined in honey bees [30]. We previously tested the bioavailability and gut uptake of these compounds in adult honey bees [57]. In this experiment, we included these PSMs to test if they are passed on to the brood and to compare bee concentrations of the added PSMs that are produced by a few plant species to concentrations of flavonoids and phenolic acids, which are omnipresent in pollen and nectar. Interestingly, all artificially added PSMs had lower adult bee concentrations than flavonoids and phenolic acids, even though they were consumed in significantly higher concentrations.

3.3. Plant secondary metabolites from pollen recovered in honey bee larvae

Larval PSM diversity was significantly reduced compared with the pollen diet, except for the phenolic acids (figure 2). All phenolic acids with a pollen abundance ≥5% were recovered in the larvae (figures 1 and 2b ). The PCA showed that larvae have distinct PSM profiles primarily composed of a few flavonoid aglycones and a mix of phenolic acids (figure 3), differing from pollen PSM profiles (figure 3). None of the pollen flavonoid glycosides were detected in the larvae, and medicarpin and quercetin were the only flavonoids consistently present across all three instars (figures 1 and 2; electronic supplementary material, table S12). Kaempferol was also detected in the fifth instar larvae and in the fourth instar larvae from hive 1 (figure 1). It is unclear if larvae can perform de-glycosylations. In this experiment, the jelly was not analysed but work by others has identified flavonoid glycosides in jelly [37], pointing to aglycones being formed in the larvae and not prior to the transmission into nurse bee jelly. The most abundant phenolic acids in the third instar were cinnamic, salicylic, 4-hydroxybenzoic, ferulic, p-coumaric and caffeic acids (figure 1 and electronic supplementary material, table S12). In the fourth and fifth instar, the abundance of caffeic and p-coumaric acids increased, making these the main phenolic acids along with ferulic acid (figure 1). Of the six compounds artificially added to the pollen diet, only gelsemine, atropine, amygdalin and senkirkine were recovered in the larvae (figure 1). Methyllycaconitine and triptolide, which had the lowest concentration in the diet, were not detected. This suggests that compound concentrations in the nurse bees’ diet influence PSM transmission to the brood, but neither the artificially added PSMs nor the clover pollen phenolic acids and flavonoids had a linear relationship with larval PSM concentrations. For all PSM groups, the r 2 value was between 0.1 and 0.3 regardless of larva instar (electronic supplementary material, figure S6). PSM concentrations differed between larvae collected from the three miniature hives (figure 1). We attribute this to variations in nurse bee food intake between the hives (total intake day 1–5: H1, 24 mg per bee; H2, 21 mg per bee; H3, 25 mg per bee), unequal numbers of larvae, and differences in jelly quantities provisioned by the nurses. Despite differences in absolute concentrations, the per cent distribution across each PSM group was similar across the hives (figures 1 and 2).

In third instar larvae, pollen grains were absent, except for one larva, which contained two pollen grains. Fourth instar larvae contained 1–3 pollen grains while the fifth instar had 10–20 clover pollen grains per larvae. Representative images from the analyses are included in electronic supplementary material, figures S2–S5. Since larvae cannot produce PSMs and the third instar larvae were pollen-free, we conclude that the PSMs detected in the early instar originate from jelly provisioned by nurses. The diverse PSM profile further documents that various pollen PSMs are indeed passed to the brood via the jelly. However, the significantly reduced larval PSM diversity (figure 2) suggests that some dietary components are not transmitted. PSMs in later instars may originate from either jelly or pollen as small amounts of pollen grains were identified in the guts of fourth and fifth instar larvae. Given the diversity of compounds detected in the pollen-free third instar larvae and the few pollen grains found in the older larvae, we conclude that jelly is the main PSM transmission pathway.

Transmission of dietary components into nurse bee jelly has only been studied for a few xenobiotic compounds. The concentrations of the pyrrolizidine alkaloid echimidine in jelly was reduced by three orders of magnitude compared with the pollen ingested by the nurses [36]. Similarly, several pesticides show reduced concentrations in jelly, while others seem to be absent despite being consumed by the nurses [79]. These observations suggest that xenobiotic transmission into jelly is highly compound-dependent and that jelly xenobiotic concentrations are significantly lower than in the nurse bees’ diet. This has led to the hypothesis that nurse bees feeding larvae jelly instead of pollen and nectar serves as a protective mechanism, reducing brood exposure to potentially toxic compounds [36]. Toxicity is rarely assessed in both adult bees and brood but larvae appear to be more sensitive to potentially toxic PSMs [36], and the activity of cytochrome P450s, mediating xenobiotic detoxification, is low in early instar larvae [80]. However, differences in toxic effects between adult bees and larvae may vary significantly depending on the compound. For example, larvae are less tolerant than adult bees to the pyrrolizidine alkaloid echimidine [36] but not to nicotine [81]. In this study, we found significantly reduced larval diversity of flavonoids and the PSMs artificially added to the diet. These results support the notion of reduced transmission of these PSMs into the jelly, thereby minimizing larval exposure to some dietary components. In contrast, all dietary phenolic acids were recovered in the larvae, suggesting that these PSMs are generally passed. Flavonoids and phenolic acids were also more abundant in the larvae than the PSMs artificially added to the diet, even though the latter had the highest diet concentration (figure 1). These results do not support a general protective effect of nurses feeding the larvae jelly that reduces larval concentrations across all PSMs but rather point to a highly compound-dependent PSM exposure of the brood. We did not measure the jelly PSM composition, so we cannot rule out that the discrepancies in PSM diversity between larvae and pollen diet might also be driven by differences in larval metabolization of individual PSMs or a combination of both. The ability of honey bee larvae to metabolize PSMs is poorly understood, but nicotine degradation was recently documented in late instar larvae [81], while early instars have not been tested.

Interestingly, our results showed that the concentration of several PSMs significantly increased from the third to the fifth larval instar. Notable increases in FC were observed for quercetin (FC 5–9), caffeic acid (FC 12–400), p-coumaric acid (FC 4–12), atropine (FC 2–14), gelsemine (FC 4–28), amygdalin (FC 10–17) and senkirkine (FC 2–16) (numbers in parentheses indicate FC ranges across H1, H2, H3; see figure 5 and electronic supplementary material, table S8). Although FC values varied among individual hives, the trends were consistent, particularly between the third and the fifth instars (figure 5). These results suggest that some PSMs accumulate in larvae as they develop. Few PSMs showed negative FC, indicating a decrease in concentration, and these decreases were generally non-significant (figure 5). During larval development, PSM depletion can only happen via metabolic degradation, as the midgut is not connected to the hindgut until the pre-pupal stage, causing undigested material to accumulate in the midgut [82]. We hypothesize that some PSMs increase in concentration because the larvae either metabolize these compounds very slowly or not at all, leading to accumulation across instars. Continuous PSM exposure could eventually lead to larval concentrations exceeding thresholds for acute toxicity or sub-lethal negative impacts on larval health. The PSMs artificially added to the pollen diet can be toxic to adult bees [5,54] and all increased with larval age (atropine, gelsemine, amygdalin and senkirkine). Toxic effects are highly concentration-dependent and exposure to sub-lethal levels may even be health-promoting. For example, amygdalin has an adult honey bee LD50 of 0.003% [5] but can reduce virus titres [83] and trypanosomatid parasite Lotmaria passim infections [83] at lower concentrations. Thus, the accumulation of dietary PSMs in larvae may alter bioactivities and potentially lead to detrimental toxic effects. When honey bees have access to a diverse range of forage plants, the consumption of individual potentially toxic PSMs is likely diminished as the bees mix pollen and nectar from several plant species with different PSM compositions. However, if they continuously forage on plants containing toxic substances, e.g. when plant species diversity is low, this increases exposure to potentially problematic dietary components. This scenario may pose a higher risk to the brood if some PSMs accumulate. Negative impacts will depend on larval toxicity thresholds, for which virtually no data exist. Our results emphasize the need to determine larval PSM tolerances, especially considering changing floral resources due to season, weather conditions, changes in land use, climatic shifts or transhumance of colonies by beekeepers, all of which could significantly alter PSM exposure.

Figure 5. Changes in plant secondary metabolite (PSM) concentrations with larval age. The concentrations of several PSMs increase significantly with larval age from the third to the fifth instar. The right side of the heatmap shows fold-change (FC) values for each miniature hive (H1, H2, H3), indicating changes in larval concentrations between instars. The left side displays the p-values from one-way ANOVAs used to determine if the concentration changes within each hive were significant. PSMs that were not detected are labelled with an ‘x’. FC and p-values are provided in electronic supplementary material, tables S8 and S9, while the statistical outputs are in electronic supplementary material, table S10.

Changes in PSM concentrations with larvae age.

Flavonoids and phenolic acids were the most abundant PSMs in the larvae. Unlike the artificially added PSMs, these compounds appear to have low toxic potential and are associated with health-promoting activities in adult honey bees [10,15,23]. Dominant larval PSMs like caffeic acid, p-coumaric acid and quercetin can promote metabolism [25,84] and increase the LD50 values of some pesticides in adult honey bees [23]. Although the influence of these compounds on pesticide detoxification in larvae is not yet determined, Mao et al. [80] found that quercetin feeding upregulated the expression of the same P450 genes in both larvae and adult honey bees, suggesting similar impacts on pesticide detoxification. Numerous studies have reported pesticide contamination of honey bee bread and wax combs [85–88], documenting pesticide exposure of the brood. Honey bees reared in pesticide-contaminated combs experience delayed larval development, delayed emergence and reduced longevity [89]. Honey bee nurses feeding on pollen that supplies the larvae with biologically adequate levels of p-coumaric acid, quercetin, and potentially other structurally related compounds may mitigate these effects by enhancing brood pesticide tolerance and detoxification. Flavonoids and phenolic acids also impact other aspects of honey bee health [18,90,91], such as upregulating innate immunity genes and genes encoding for the production of antimicrobial peptides in both adults and larvae [14,15]. The concentration dependence of these activities is not yet determined, and the increasing concentrations of caffeic acid, p-coumaric acid and quercetin with larval age documented by this study may alter their effects from health promoting to harmful. For example, the accumulation of quercetin in adult honey bees exposed to a triazole fungicide that inhibits quercetin metabolism impaired the bees’ ATP energy production [80].

3.4. Comparison of plant secondary metabolite profiles between adult honey bees and larvae

Larval PSM diversity was significantly reduced compared with adult bees for all compound groups except phenolic acids, with an abundance above 5% (figure 2). These differences were mainly driven by flavonoids and the two artificially added compounds, triptolide and methyllycaconitine, which were absent in larvae and had low concentrations in adult bees. The PCA showed a clear separation between larval and adult bee PSM profiles. Phenolic acids were dominant in larvae, while adult bees contained higher levels of flavonoids and the PSMs artificially added to the diet (figure 6). The differences between adult bees and larvae are likely due to variations in the transmission of dietary PSMs to the brood and differences in PSM metabolism between developmental stages. We consistently detected lower concentrations of some of the flavonoids and all artificially added PSMs in larvae compared with adult bees (figures 1 and 6), supporting the proposed protective effect of nurses feeding larvae jelly to reduce larval PSM exposure (discussed in §3.3). However, several phenolic acids had the highest concentration in larvae, and some larval samples also exceeded adult bee concentrations of medicarpin and quercetin (electronic supplementary material, tables S11 and S12). These observations suggest that PSM transmission via jelly is highly compound-dependent rather than indicating a general protective effect of honey bee nursing.

Figure 6. Comparison of plant secondary metabolite (PSM) profiles of adult honey bees and larvae. The PCA scores plot (a) shows distinct PSM profiles of adult bees and larvae (bees, n = 5 pools; larvae, n = 3 pools per hive and instar). The loadings plot (b) shows that larval stages are predominantly characterized by phenolic acids, whereas adult bees exhibit higher concentrations of flavonoids and the PSMs that were artificially added to the diet. Numbers correspond to individual PSMs: 1, astragalin; 2, hyperoside; 3, isoquercitin; 4, kaempferol; 5, luteolin; 6, medicarpin; 7, naringenin; 8, pratensein; 9, quercetin; 10, quercetin-xyl-gal; 11, 4-hydroxybenzoic acid; 12, caffeic acid; 13, chlorogenic acid; 14, cinnamic acid; 15, coumaric acid; 16, ferulic acid; 17, gallic acid; 18, protocatechuic acid; 19, salicylic acid; 20, sinapic acid; 21, syringic acid; 22, vanillic acid; 23, amygdalin; 24, atropine; 25, gelsemine; 25, methyllycaconitine; 26, senkirkine; 27, triptolide.

Comparison of PSM profiles of adult bees and larvae.

The flavonoid ring-system degradation products, vanillic and protocatechuic acids, detected in high concentrations in adult bees were only minor components in larvae (figure 1). This suggests different larval flavonoid metabolization pathways or an inability of larvae to metabolize the flavonoid ring system. This is consistent with the observed increase in flavonoid aglycones from the third to the fifth larval instar (figure 5). Differences in xenobiotic metabolism between adult bees and larvae remain largely unexplored, especially for PSM xenobiotics. However, metabolization of the agrochemical formulant N-methyl-2-pyrrolidone was faster in adult bees than in larvae, which also formed different metabolites from adult bees [92]. Pathway and rate differences between larvae and adult bees likely reflect changes in enzyme activity documented across developmental stages [92]. Each developmental stage of honey bees also has a distinct gut microbial composition [93,94], with larval guts possessing few or no bacteria [95]. Since gut microbiota plays a role in degrading flavonoids [29] and other dietary PSMs [30], the near-absence of gut microbiota in larval stages may contribute to the substantial decrease in flavonoid ring-system metabolites compared with adult bees documented by our study.

4. Conclusion

This experiment presents the first documentation of pollen diet PSMs from multiple compound classes in honey bee larvae and the first comparison of diet PSMs across honey bee developmental stages. Our results reveal distinct PSM profiles of adult honey bees and three larval stages, differing from their pollen diet composition. These differences are attributed to post-ingestion PSM metabolism and variations in the transmission of individual PSMs to the brood via nurse bee jelly. Both bees and larvae had the highest concentrations of phenolic acids and flavonoid aglycones, which are naturally present in the clover pollen diet and omnipresent in pollen and nectar. In contrast, the alkaloids, cyanogenic glycoside and diterpene produced by fewer plant species were less abundant in both adult bees and larvae despite being consumed in higher concentrations.

Our findings propose several new hypotheses about PSM fate in adult honey bees and their larvae, with potential implications for honey bee health. Notably, our results point to differences in flavonoid metabolism between adult bees and larvae and that PSM transmission to the brood via nurse bee jelly is highly compound-dependent. The increasing concentrations of several PSMs with larval age suggest a reduced ability of honey bee larvae to degrade certain dietary PSMs, leading to accumulation and potential negative effects on brood health. Future experiments should focus on these hypotheses to improve our understanding of PSM metabolism and the impacts of dietary PSMs on honey bee health.

Acknowledgements

We thank Betinna Woller for supplying the honey bee hives, Tove Steenberg and Claus Weirsøe Dahl for performing palynological analyses of the pollen used for the diet, and Stine Krogh Steffensen, who had developed the HPLC-MS/MS method for phenolic acids for another project and kindly allowed us to use it for this work.

Ethics

These experiments were conducted in Denmark where ethical approval and permits can currently only be obtained for experiments involving mammals. In Denmark, there are no guidelines and no system in place for handling ethical considerations regarding research with honey bees or other invertebrates. Thus, we could neither apply for nor obtain ethical approval for these experiments. However, we have taken great care to conduct the experiments in an ethically responsible way. As detailed in the Material and methods section, the bees were kept according to the guidelines for standard methods for Apis mellifera research (published collectively as the COLOSS Beebook; relevant sections referenced in the paper). In brief, we kept the small bee colonies in standard miniature hives in our apiary and fed them with a natural pollen diet fortified with below-toxicity concentrations of plant compounds that are an integrated part of the bees’ natural diet. At the end of the experiments, the colonies were knocked out with CO2 and sacrificed by freezing.

Data accessibility

The HPLC-MS/MS data tables are included as electronic supplementary material [96]. The raw HPLC-MS/MS data files are available via Dryad [97].

Declaration of AI use

We have not used AI-assisted technologies in creating this article.

Authors’ contributions

N.H.V.: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resources, validation, visualization, writing—original draft, writing—review and editing; B.B.L.: investigation, validation; P.K.: conceptualization, investigation, methodology, resources, writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

This work was supported by a grant from the Independent Research Fund Denmark for N.H.V. (grant number DFF-6111-00441).
==== Refs
References

1. Bennett RN , Wallsgrove RM . 1994 Secondary metabolites in plant defence mechanisms. New Phytol. 127 , 617–633. (10.1111/j.1469-8137.1994.tb02968.x)33874382
2. Erb M , Kliebenstein DJ . 2020 Plant secondary metabolites as defenses, regulators, and primary metabolites: the blurred functional trichotomy. Plant Physiol. 184 , 39–52. (10.1104/pp.20.00433)32636341
3. Bankova V , Popova M , Trusheva B . 2018 The phytochemistry of the honeybee. Phytochemistry 155 , 1–11. (10.1016/j.phytochem.2018.07.007)30053651
4. Stevenson PC . 2020 For antagonists and mutualists: the paradox of insect toxic secondary metabolites in nectar and pollen. Phytochem. Rev. 19 , 603–614. (10.1007/s11101-019-09642-y)
5. Detzel A , Wink M . 1993 Attraction, deterrence or intoxication of bees (Apis mellifera) by plant allelochemicals. Chemoecology 4 , 8–18. (10.1007/BF01245891)
6. Irwin RE , Cook D , Richardson LL , Manson JS , Gardner DR . 2014 Secondary compounds in floral rewards of toxic rangeland plants: impacts on pollinators. J. Agric. Food Chem. 62 , 7335–7344. (10.1021/jf500521w)24766254
7. Parachnowitsch AL , Manson JS , Sletvold N . 2019 Evolutionary ecology of nectar. Ann. Bot. 123 , 247–261. (10.1093/aob/mcy132)30032269
8. Negri P et al . 2019 Towards precision nutrition: a novel concept linking phytochemicals, immune response and honey bee health. Insects 10 , 401. (10.3390/insects10110401)31726686
9. Richardson LL , Adler LS , Leonard AS , Andicoechea J , Regan KH , Anthony WE , Manson JS , Irwin RE . 2015 Secondary metabolites in floral nectar reduce parasite infections in bumblebees. Proc. R. Soc. B 282 , 20142471. (10.1098/rspb.2014.2471)
10. Berenbaum MR , Calla B . 2021 Honey as a functional food for Apis mellifera. Annu. Rev. Entomol. 66 , 185–208. (10.1146/annurev-ento-040320-074933)32806934
11. Aurori AC , Bobiş O , Dezmirean DS , Mărghitaş LA , Erler S. 2016 Bay laurel (Laurus nobilis) as potential antiviral treatment in naturally BQCV infected honeybees. Virus Res. 222 , 29-33. (10.1016/j.virusres.2016.2016.05.024)27235809
12. Hsieh EM , Berenbaum MR , Dolezal AG . 2020 Ameliorative effects of phytochemical ingestion on viral infection in honey bees. Insects 11 , 698–715. (10.3390/insects11100698)33066263
13. Negri P , Maggi MD , Ramirez L , De Feudis L , Szwarski N , Quintana S , Eguaras MJ , Lamattina L . 2015 Abscisic acid enhances the immune response in Apis mellifera and contributes to the colony fitness. Apidologie 46 , 542–557. (10.1007/s13592-014-0345-7)
14. Palmer-Young EC , Tozkar CÖ , Schwarz RS , Chen Y , Irwin RE , Adler LS , Evans JD . 2017 Nectar and pollen phytochemicals stimulate honey bee (Hymenoptera: Apidae) immunity to viral infection. J. Econ. Entomol. 110 , 1959–1972. (10.1093/jee/tox193)28981688
15. Mao W , Schuler MA , Berenbaum MR . 2013 Honey constituents up-regulate detoxification and immunity genes in the western honey bee Apis mellifera. Proc. Natl Acad. Sci. USA 110 , 8842–8846. (10.1073/pnas.1303884110)23630255
16. Maistrello L , Lodesani M , Costa C , Leonardi F , Marani G , Caldon M , Mutinelli F , Granato A . 2008 Screening of natural compounds for the control of nosema disease in honeybees (Apis mellifera). Apidologie 39 , 436–445. (10.1051/apido:2008022)
17. Costa C , Lodesani M , Maistrello L . 2010 Effect of thymol and resveratrol administered with candy or syrup on the development of Nosema ceranae and on the longevity of honeybees (Apis mellifera L.) in laboratory conditions. Apidologie 41 , 141–150. (10.1371/journal.pone.0144668)
18. Bernklau E , Bjostad L , Hogeboom A , Carlisle A , Arathi HS . 2019 Dietary phytochemicals, honey bee longevity and pathogen tolerance. Insects 10 , 14–25. (10.3390/insects10010014)30626025
19. Gherman BI , Denner A , Bobiş O , Dezmirean DS , Mărghitaş LA , Schlüns H , Moritz RFA , Erler S . 2014 Pathogen-associated self-medication behavior in the honeybee Apis mellifera. Behav. Ecol. Sociobiol. 68 , 1777–1784. (10.1007/s00265-014-1786-8)
20. Szawarski N et al . 2020 Short communication: antimicrobial activity of indoleacetic, gibberellic and coumaric acids against Paenibacillus larvae and its toxicity against Apis mellifera. Span. J. Agric. Res. 18 , e05SC01. (10.5424/sjar/2020181-15158)
21. Fitch G , Figueroa LL , Koch H , Stevenson PC , Adler LS . 2022 Understanding effects of floral products on bee parasites: mechanisms, synergism, and ecological complexity. Int. J. Parasitol. Parasites Wildl. 17 , 244–256. (10.1016/j.ijppaw.2022.02.011)35299588
22. Geldert C , Abdo Z , Stewart JE , Arathi HS . 2021 Dietary supplementation with phytochemicals improves diversity and abundance of honey bee gut microbiota. J. Appl. Microbiol. 130 , 1705–1720. (10.1111/jam.14897)33058297
23. Johnson RM , Mao W , Pollock HS , Niu G , Schuler MA , Berenbaum MR . 2012 Ecologically appropriate xenobiotics induce cytochrome p450s in Apis mellifera. PLoS One 7 , e31051. (10.1371/journal.pone.0031051)22319603
24. Arathi HS , Bernklau E . 2021 Context-dependent effect of dietary phytochemicals on honey bees exposed to a pesticide, thiamethoxam. J. Insect Sci. 21 , 11. (10.1093/jisesa/ieab053)
25. Ardalani H , Vidkjær NH , Laursen BB , Kryger P , Fomsgaard IS . 2021 Dietary quercetin impacts the concentration of pesticides in honey bees. Chemosphere 262 , 127848. (10.1016/j.chemosphere.2020.127848)32771708
26. Ardalani H , Vidkjær NH , Kryger P , Fiehn O , Fomsgaard IS . 2021 Metabolomics unveils the influence of dietary phytochemicals on residual pesticide concentrations in honey bees. Environ. Int. 152 , 106503. (10.1016/j.envint.2021.106503)33756430
27. Berenbaum MR , Johnson RM . 2015 Xenobiotic detoxification pathways in honey bees. Curr. Opin. Insect Sci. 10 , 51–58. (10.1016/j.cois.2015.03.005)29588014
28. Gong Y , Diao Q . 2017 Current knowledge of detoxification mechanisms of xenobiotic in honey bees. Ecotoxicology 26 , 1–12. (10.1007/s10646-016-1742-7)27819118
29. Kešnerová L , Mars RAT , Ellegaard KM , Troilo M , Sauer U , Engel P . 2017 Disentangling metabolic functions of bacteria in the honey bee gut. PLoS Biol. 15 , e2003467. (10.1371/journal.pbio.2003467)29232373
30. Motta EVS , Gage A , Smith TE , Blake KJ , Kwong WK , Riddington IM , Moran N , Koch H . 2022 Host-microbiome metabolism of a plant toxin in bees. eLife 11 , e82595. (10.7554/eLife.82595)36472498
31. Luca SV , Macovei I , Bujor A , Miron A , Skalicka-Woźniak K , Aprotosoaie AC , Trifan A . 2020 Bioactivity of dietary polyphenols: the role of metabolites. Crit. Rev. Food Sci. Nutr. 60 , 626–659. (10.1080/10408398.2018.1546669)30614249
32. Koch H , Welcome V , Kendal-Smith A , Thursfield L , Farrell IW , Langat MK , Brown MJF , Stevenson PC . 2022 Host and gut microbiome modulate the antiparasitic activity of nectar metabolites in a bumblebee pollinator. Phil. Trans. R. Soc. B 377 , 20210162. (10.1098/rstb.2021.0162)35491601
33. Haydak MH . 1970 Honey bee nutrition. Annu. Rev. Entomol. 15 , 143–156. (10.1146/annurev.en.15.010170.001043)
34. Wang Y , Ma L , Zhang W , Cui X , Wang H , Xu B . 2016 Comparison of the nutrient composition of royal jelly and worker jelly of honey bees (Apis mellifera). Apidologie (Celle) 47 , 48–56. (10.1007/s13592-015-0374-x)
35. Herbert EW , Shimanuki H . 1978 Chemical composition and nutritive value of bee-collected and bee-stored pollen. Apidologie 9 , 33–40. (10.1051/apido:19780103)
36. Lucchetti MA , Kilchenmann V , Glauser G , Praz C , Kast C . 2018 Nursing protects honeybee larvae from secondary metabolites of pollen. Proc. R. Soc. B 285 , 20172849. (10.1098/rspb.2017.2849)
37. López-Gutiérrez N , Aguilera-Luiz MDM , Romero-González R , Vidal JLM , Garrido Frenich A . 2014 Fast analysis of polyphenols in royal jelly products using automated turboflowTM-liquid chromatography–orbitrap high resolution mass spectrometry. J. Chromatogr. B 973 , 17–28. (10.1016/j.jchromb.2014.09.038)
38. Büchler R et al . 2013 Standard methods for rearing and selection of Apis mellifera queens. J. Apic. Res. 52 , 1–30. (10.3896/IBRA.1.52.1.07)
39. Williams GR et al . 2013 Standard methods for maintaining adult Apis mellifera in cages under in vitro laboratory conditions. J. Apic. Res. 52 , 1–36. (10.3896/IBRA.1.52.1.04)
40. Carlsen SCK , Pedersen HA , Spliid NH , Fomsgaard IS . 2012 Fate in soil of flavonoids released from white clover (Trifolium repens L.). Appl. Environ. Soil Sci. 2012 , 1–10. (10.1155/2012/743413)
41. Carlsen SCK , Understrup A , Fomsgaard IS , Mortensen AG , Ravnskov S . 2008 Flavonoids in roots of white clover: interaction of arbuscular mycorrhizal fungi and a pathogenic fungus. Plant Soil 302 , 33–43. (10.1007/s11104-007-9452-9)
42. Carlsen SCK , Mortensen AG , Oleszek W , Piacente S , Stochmal A , Fomsgaard IS . 2008 Variation in flavonoids in leaves, stems and flowers of white clover cultivars. Nat. Prod. Commun. 3 , 1934578X0800300. (10.1177/1934578X0800300813)
43. Carlsen SCK , Fomsgaard IS . 2008 Biologically active secondary metabolites in white clover (Trifolium repens L.): a review focusing on contents in the plant, plant–pest interactions and transformation. Chemoecology 18 , 129–170. (10.1007/s00049-008-0402-7)
44. Kicel A , Wolbiś M . 2012 Study on the phenolic constituents of the flowers and leaves of Trifolium repens L. Nat. Prod. Res. 26 , 2050–2054. (10.1080/14786419.2011.637217)22117193
45. Kolodziejczyk-Czepas J . 2012 Trifolium species-derived substances and extracts: biological activity and prospects for medicinal applications. J. Ethnopharmacol. 143 , 14–23. (10.1016/j.jep.2012.06.048)22771317
46. Manson JS , Otterstatter MC , Thomson JD . 2010 Consumption of a nectar alkaloid reduces pathogen load in bumble bees. Oecologia 162 , 81–89. (10.1007/s00442-009-1431-9)19711104
47. Martinello M , Cristofoli C , Gallina A , Mutinelli F . 2014 Easy and rapid method for the quantitative determination of pyrrolizidine alkaloids in honey by ultra performance liquid chromatography-mass spectrometry: an evaluation in commercial honey. Food Control 37 , 146–152. (10.1016/j.foodcont.2013.09.037)
48. Boros B , Farkas Á , Jakabová S , Bacskay I , Kilár F , Felinger A . 2010 LC-MS quantitative determination of atropine and scopolamine in the floral nectar of Datura species. Chromatographia 71 , 43–49. (10.1365/s10337-010-1524-y)
49. Richardson LL , Bowers MD , Irwin RE . 2016 Nectar chemistry mediates the behavior of parasitized bees: consequences for plant fitness. Ecology 97 , 325–337. (10.1890/15-0263.1)27145608
50. London-Shafir I , Shafir S , Eisikowitch D . 2003 Amygdalin in almond nectar and pollen: facts and possible roles. Plant Syst. Evol. 238 , 87–95. (10.1007/s00606-003-0272-y)
51. Tan K , Guo YH , Nicolson SW , Radloff SE , Song QS , Hepburn HR . 2007 Honeybee (Apis cerana) foraging responses to the toxic honey of Tripterygium hypoglaucum (Celastraceae): changing threshold of nectar acceptability. J. Chem. Ecol. 33 , 2209–2217. (10.1007/s10886-007-9384-0)18058178
52. Cook D , Manson JS , Gardner DR , Welch KD , Irwin RE . 2013 Norditerpene alkaloid concentrations in tissues and floral rewards of larkspurs and impacts on pollinators. Biochem. Syst. Ecol. 48 , 123–131. (10.1016/j.bse.2012.11.015)
53. Kempf M , Heil S , Hasslauer I , Schmidt L , von der Ohe K , Theuring C , Reinhard A , Schreier P , Beuerle T . 2010 Pyrrolizidine alkaloids in pollen and pollen products. Mol. Nutr. Food Res. 54 , 292–300. (10.1002/mnfr.200900289)20013884
54. Reinhard A , Janke M , von der Ohe W , Kempf M , Theuring C , Hartmann T , Schreier P , Beuerle T . 2009 Feeding deterrence and detrimental effects of pyrrolizidine alkaloids fed to honey bees (Apis mellifera). J. Chem. Ecol. 35 , 1086–1095. (10.1007/s10886-009-9690-9)19777310
55. Lecocq A , Green AA , Pinheiro De Castro ÉC , Olsen CE , Jensen AB , Zagrobelny M . 2018 Honeybees tolerate cyanogenic glucosides from clover nectar and flowers. Insects 9 , 31. (10.3390/insects9010031)29534004
56. Singaravelan N , Nee’man G , Inbar M , Izhaki I . 2005 Feeding responses of free-flying honeybees to secondary compounds mimicking floral nectars. J. Chem. Ecol. 31 , 2791–2804. (10.1007/s10886-005-8394-z)16365705
57. Vidkjær NH , Fomsgaard IS , Kryger P . 2021 LC-MS/MS quantification reveals ample gut uptake and metabolization of dietary phytochemicals in honey bees (Apis mellifera). J. Agric. Food Chem. 69 , 627–637. (10.1021/acs.jafc.0c03584)33416324
58. Nešović M , Gašić U , Tosti T , Horvacki N , Šikoparija B , Nedić N , Blagojević S , Ignjatović L , Tešić Ž . 2020 Polyphenol profile of buckwheat honey, nectar and pollen. R. Soc. Open Sci. 7 , 201576. (10.1098/rsos.201576)33489289
59. Kicel A , Wolbis M . 2006 Phenolic acids in flowers and leaves of Trifolium repens L. and Trifolium pratense L. Herba. Pol. 54 , 52–60.
60. Zhou J , Qi Y , Ritho J , Zhang Y , Zheng X , Wu L , Li Y , Sun L . 2015 Flavonoid glycosides as floral origin markers to discriminate of unifloral bee pollen by LC–MS/MS. Food Control 57 , 54–61. (10.1016/j.foodcont.2015.03.035)
61. Sobral F , Calhelha RC , Barros L , Dueñas M , Tomás A , Santos-Buelga C , Vilas-Boas M , Ferreira ICFR . 2017 Flavonoid composition and antitumor activity of bee bread collected in northeast Portugal. Molecules 22 , 248–259. (10.3390/molecules22020248)28178217
62. Rasool S , Vidkjær NH , Hooshmand K , Jensen B , Fomsgaard IS , Meyling NV . 2021 Seed inoculations with entomopathogenic fungi affect aphid populations coinciding with modulation of plant secondary metabolite profiles across plant families. New Phytol. 229 , 1715–1727. (10.1111/nph.16979)33006149
63. Schittko U , Burghardt F , Fiedler K , Wray V , Proksch P . 1999 Sequestration and distribution of flavonoids in the common blue butterfly Polyommatus icarus reared on Trifolium repens. Phytochemistry 51 , 609–614. (10.1016/S0031-9422(98)00746-8)
64. Ferreres F , Fernandes F , Pereira DM , Pereira JA , Valentão P , Andrade PB . 2009 Phenolics metabolism in insects: Pieris brassicae-Brassica oleracea var. costata ecological duo. J. Agric. Food Chem. 57 , 9035–9043. (10.1021/jf901538j)19764727
65. Fu X , Chai CL , Li YP , Li P , Luo SH , Li Q , Li MW , Liu YQ . 2021 Metabolomics reveals abundant flavonoids in edible insect Antheraea pernyi. J. Asia Pac. Entomol. 24 , 711–715. (10.1016/j.aspen.2021.06.004)
66. Geuder M , Wray V , Fiedler K , Proksch P . 1997 Sequestration and metabolism of host-plant flavonoids by the lycaenid butterfly Polyommatus bellargus. J. Chem. Ecol. 23 , 1361–1372. (10.1023/B:JOEC.0000006469.00557.69)
67. Mizokami H , Yoshitama K . 2009 Sequestration and metabolism of host–plant flavonoids by the pale grass blue, Pseudozizeeria maha (Lepidoptera: Lycaenidae). Entomol. Sci. 12 , 171–176. (10.1111/j.1479-8298.2009.00322.x)
68. Salminen JP , Lahtinen M , Lempa K , Kapari L , Haukioja E , Pihlaja K . 2004 Metabolic modifications of birch leaf phenolics by an herbivorous insect: detoxification of flavonoid aglycones via glycosylation. Z. Naturforsch. C 59 , 437–444. (10.1515/znc-2004-5-627)18998416
69. Liu H , Huo X , Ding L , Feng X , Jiang M , Pan G , Chen L , Qiu F . 2016 Metabolic profiling of luteolin‐7‐O‐glucoside in rat urine, plasma, bile and feces after oral administration using ultra‐high‐performance liquid chromatography/quadrupole time‐of‐flight mass spectrometry. Rapid Commun. Mass Spectrom. 30 , 447–459. (10.1002/rcm.7456)
70. Braune A , Blaut M . 2016 Bacterial species involved in the conversion of dietary flavonoids in the human gut. Gut Microbes 7 , 216–234. (10.1080/19490976.2016.1158395)26963713
71. Feng X , Li Y , Brobbey Oppong M , Qiu F . 2018 Insights into the intestinal bacterial metabolism of flavonoids and the bioactivities of their microbe-derived ring cleavage metabolites. Drug Metab. Rev. 50 , 343–356. (10.1080/03602532.2018.1485691)30010437
72. Selma MV , Espín JC , Tomás-Barberán FA . 2009 Interaction between phenolics and gut microbiota: role in human health. J. Agric. Food Chem. 57 , 6485–6501. (10.1021/jf902107d)19580283
73. Chesson A . 1999 Hydroxycinnamic acids in the digestive tract of livestock and humans. J. Sci. Food Agric. 79 , 373. (10.1002/(sici)1097-0010(19990301)79:33.0.co;2-6)
74. Pontoh J , Low NH . 2002 Purification and characterization of β-glucosidase from honey bees (Apis mellifera). Insect Biochem. Mol. Biol. 32 , 679–690. (10.1016/S0965-1748(01)00147-3)12020842
75. Ricigliano VA , Fitz W , Copeland DC , Mott BM , Maes P , Floyd AS , Dockstader A , Anderson KE . 2017 The impact of pollen consumption on honey bee (Apis mellifera) digestive physiology and carbohydrate metabolism. Arch. Insect Biochem. Physiol. 96 , e21406. (10.1002/arch.21406)
76. da Silva RR , Moraes MM , Camara CAG , Ramos CS . 2015 Change in the chemical profile of Mangifera indica leaves after their metabolism in the Tropidacris collaris grasshopper. Nat. Prod. Commun. 10 , 1809–1810. (10.1177/1934578X1501001104)26749802
77. Nicolson SW . 2022 Sweet solutions: nectar chemistry and quality. Phil. Trans. R. Soc. B 377 , 20210163. (10.1098/rstb.2021.0163)35491604
78. Mao W , Rupasinghe SG , Johnson RM , Zangerl AR , Schuler MA , Berenbaum MR . 2009 Quercetin-metabolizing CYP6AS enzymes of the pollinator Apis mellifera (Hymenoptera: Apidae). Comp. Biochem. Physiol. B 154 , 427–434. (10.1016/j.cbpb.2009.08.008)19737624
79. Böhme F , Bischoff G , Zebitz CPW , Rosenkranz P , Wallner K . 2018 From field to food—will pesticide-contaminated pollen diet lead to a contamination of royal jelly? Apidologie 49 , 112–119. (10.1007/s13592-017-0533-3)
80. Mao W , Schuler MA , Berenbaum MR . 2017 Disruption of quercetin metabolism by fungicide affects energy production in honey bees (Apis mellifera). Proc. Natl Acad. Sci. USA 114 , 2538–2543. (10.1073/pnas.1614864114)28193870
81. du Rand EE , Human H , Smit S , Beukes M , Apostolides Z , Nicolson SW , Pirk CWW . 2017 Proteomic and metabolomic analysis reveals rapid and extensive nicotine detoxification ability in honey bee larvae. Insect Biochem. Mol. Biol. 82 , 41–51. (10.1016/j.ibmb.2017.01.011)28161469
82. Michener CD . 2007 The bees of the world, 2nd edn. Baltimore, MD: The Johns Hopkins University Press.
83. Tauber JP , Tozkar CÖ , Schwarz RS , Lopez D , Irwin RE , Adler LS , Evans JD . 2020 Colony-level effects of amygdalin on honeybees and their microbes. Insects 11 , 783. (10.3390/insects11110783)33187240
84. Mao W , Schuler MA , Berenbaum MR . 2011 CYP9Q-mediated detoxification of acaricides in the honey bee (Apis mellifera). Proc. Natl Acad. Sci. USA 108 , 12657–12662. (10.1073/pnas.1109535108)21775671
85. Ostiguy N , Drummond FA , Aronstein K , Eitzer B , Ellis JD , Spivak M , Sheppard WS . 2019 Honey bee exposure to pesticides: a four-year nationwide study. Insects 10 , 13–46. (10.3390/insects10010013)30626027
86. Mullin CA , Frazier M , Frazier JL , Ashcraft S , Simonds R , vanEngelsdorp D , Pettis JS . 2010 High levels of miticides and agrochemicals in North American apiaries: implications for honey bee health. PLoS One 5 , e9754. (10.1371/journal.pone.0009754)20333298
87. Xiao J et al . 2022 Analysis of honey bee exposure to multiple pesticide residues in the hive environment. Sci. Total Environ. 805 , 150292. (10.1016/j.scitotenv.2021.150292)34536857
88. Wilmart O , Legrève A , Scippo ML , Reybroeck W , Urbain B , de Graaf DC , Spanoghe P , Delahaut P , Saegerman C . 2021 Honey bee exposure scenarios to selected residues through contaminated beeswax. Sci. Total Environ. 772 , 145533. (10.1016/j.scitotenv.2021.145533)33770874
89. Wu JY , Anelli CM , Sheppard WS . 2011 Sub-lethal effects of pesticide residues in brood comb on worker honey bee (Apis mellifera) development and longevity. PLoS One 6 , e14720. (10.1371/journal.pone.0014720)21373182
90. Liao LH , Wu WY , Berenbaum MR . 2017 Impacts of dietary phytochemicals in the presence and absence of pesticides on longevity of honey bees (Apis mellifera). Insects 8 , 22–34. (10.3390/insects8010022)28216580
91. Wong MJ , Liao LH , Berenbaum MR . 2018 Biphasic concentration-dependent interaction between imidacloprid and dietary phytochemicals in honey bees (Apis mellifera). PLoS One 13 , e0206625. (10.1371/journal.pone.0206625)30383869
92. Fine JD , Mullin CA . 2017 Metabolism of N-methyl-2-pyrrolidone in honey bee adults and larvae: exploring age related differences in toxic effects. Environ. Sci. Technol. 51 , 11412–11422. (10.1021/acs.est.7b03291)28858486
93. Khan KA et al . 2020 Structural diversity and functional variability of gut microbial communities associated with honey bees. Microb. Pathog. 138 , 103793. (10.1016/j.micpath.2019.103793)31626917
94. Hroncova Z et al . 2015 Variation in honey bee gut microbial diversity affected by ontogenetic stage, age and geographic location. PLoS One 10 , e0118707. (10.1371/journal.pone.0118707)25768309
95. Martinson VG , Moy J , Moran NA . 2012 Establishment of characteristic gut bacteria during development of the honeybee worker. Appl. Environ. Microbiol. 78 , 2830–2840. (10.1128/AEM.07810-11)22307297
96. Vidkjær NH , Laursen BB , Kryger P . 2024 Data from: Phytochemical profiles of honey bees (Apis mellifera) and their larvae differ from the composition of their pollen diet. Figshare. (10.6084/m9.figshare.c.7454632)
97. Vidkjær NH , Laursen B , Kryger P . 2024 Phytochemical profiles of honey bees (Apis mellifera) and their larvae differ from the composition of their pollen diet. Dryad Digital Repository. (10.5061/dryad.83bk3j9w3)
