Disentangling metabolic functions of bacteria in the honey bee gut

Abstract

It is presently unclear how much individual community members contribute to the overall metabolic output of a gut microbiota. To address this question, we used the honey bee, which harbors a relatively simple and remarkably conserved gut microbiota with striking parallels to the mammalian system and importance for bee health. Using untargeted metabolomics, we profiled metabolic changes in gnotobiotic bees that were colonized with the complete microbiota reconstituted from cultured strains. We then determined the contribution of individual community members in mono-colonized bees and recapitulated our findings using in vitro cultures. Our results show that the honey bee gut microbiota utilizes a wide range of pollen-derived substrates, including flavonoids and outer pollen wall components, suggesting a key role for degradation of recalcitrant secondary plant metabolites and pollen digestion. In turn, multiple species were responsible for the accumulation of organic acids and aromatic compound degradation intermediates. Moreover, a specific gut symbiont, Bifidobacterium asteroides, stimulated the production of host hormones known to impact bee development. While we found evidence for cross-feeding interactions, approximately 80% of the identified metabolic changes were also observed in mono-colonized bees, with Lactobacilli being responsible for the largest share of the metabolic output. These results show that, despite prolonged evolutionary associations, honey bee gut bacteria can independently establish and metabolize a wide range of compounds in the gut. Our study reveals diverse bacterial functions that are likely to contribute to bee health and provide fundamental insights into how metabolic activities are partitioned within gut communities.

Fig 1. Overview of the experimental setup to characterize metabolic activities of the honey bee gut microbiota.

Newly emerged adult bees were either kept microbiota-depleted (MD), colonized with a reconstituted community of the seven predominant species of the bee gut microbiota (CL), or mono-colonized with one of the seven species separately. Bees received sterilized bee pollen and sugar water as diet. Ten days after colonization, metabolites were extracted from the honey bee guts and subjected to untargeted metabolomics to (1) reveal overall metabolic changes in CL versus MD bees and (2) identify which community member could explain these metabolic changes in the gut. As a control, we additionally analyzed 10-d-old hive bees that were colonized by the native microbiota under natural conditions in the colony (not shown in this figure). To recapitulate findings in vitro, individual community members were cultured in pollen-conditioned medium, and metabolic changes in this medium were profiled using untargeted metabolomics. MS, mass spectrometry; Q-TOF, quadrupole-time of flight.

Fig 2. Bacterial colonization levels in the guts of microbiota-depleted (MD), colonized (CL), and hive bees.

(A) Total bacterial loads in the gut of 10-d-old MD bees (n = 21), CL bees (n = 18), and hive bees (n = 16) were assessed by quantitative PCR (qPCR) with universal bacterial 16S rRNA primers. (B) The bacterial loads of the seven predominant community members used for experimental colonizations were assessed by qPCR with species-specific 16S rRNA primers for the same bees as shown in panel A. Black lines show median values. Samples with <10⁵ bacterial cells per gut are shown below the red line, which we consider the threshold of detection. Primer characteristics are summarized in S2 Table. n.s., not significant; *P < 0.05; **P < 0.01; and ***P < 0.001 (Wilcoxon Rank Sum test, Benjamini and Hochberg adjusted [BH adj.]). The numerical data can be found in S1 Data.

Fig 3. Metabolite changes between microbiota-depleted (MD) and colonized (CL) bees.

An Orthogonal Projection of Least Squares-Differentiation Analysis (OPLS-DA) based S-plot of metabolite changes shows the ions responsible for CL and MD separation. The inset shows OPLS-DA separation between CL and MD along the component that was used for correlating ion intensities. Experiment 2 data (see S2A Data) was used for this plot, and annotated ions that were not robustly significantly different between CL and MD in both experiments are plotted in grey. Ions with a first annotation belonging to an enriched category (see S3A Data) are plotted in color, except for the category “amino acids and derivatives”, which did not meet the significance threshold for enrichment but was deemed relevant. The “purine nucleosides and analogues” and “pyrimidine nucleosides and analogues” categories were combined into “nucleosides and analogs” for coloring only. The boxed areas show the m/z [M-H⁺]^- of the ion and the first annotation name of the most discriminatory ions, sorted by covariance. Asterisks indicate ions with ambiguous annotations. The numerical data can be found in S1 Data. Conjug., conjugates; Deriv., derivatives; FC, fold change; int., intensity.

Fig 4. Overview of metabolite changes explained by different community members of the bee gut microbiota.

(A) Bar graphs show the fraction of the metabolic changes explained by mono-colonizations and hive bees for substrates (240 ions) and products (132 ions). The category “Total” indicates the total number of ions explained by mono-colonizations, thus excluding hive bees. Heatmap representation of enrichment P values (one-sided Fisher’s exact test P < 0.05) are provided for compound categories enriched in one or several mono-colonizations. (B–E) Z-score transformed ion intensities of selected substrate and product ions are shown for all treatment groups. (B) Four glycosylated flavonoid substrates. (C) Two substrates from the outer pollen wall. (D) Two products corresponding to host-derived metabolites. (E) Succinate, one of the major fermentation products. Groups depicted in color highlight treatment groups displaying a significant difference compared to MD bees in the same direction as the CL versus MD difference (one-way analysis of variance [ANOVA], Tukey honest significant difference [HSD] post hoc test at 99% confidence, P ≤ 0.05). Plots for all 372 ions are provided in S8 Data. Ba, B. apis mono-colonized; Bi, B. asteroides mono-colonized; CL, colonized with the reconstituted microbiota; F4, Firm-4 mono-colonized; F5, Firm-5 mono-colonized; Fp, F. perrara mono-colonized; Ga, G. apicola mono-colonized; Hive, hive bees; MD, microbiota-depleted; Sa, S. alvi mono-colonized. The numerical results of the full enrichment analysis, bar graphs, and mono-colonization plots are provided in S3 Data, S1 Data and S8 Data, respectively.

Fig 5. Recapitulation of flavonoid degradation patterns by gut bacteria during in vitro growth in pollen-conditioned medium.

(A) Line graphs show the growth of each community member in control medium and pollen-conditioned medium based on colony-forming unit (CFU) counts at time points 0 h and 16 h. Values are the mean of five replicates, with error bars indicating standard deviation. **P < 0.01, ***P < 0.001 (Welch’s t test). Volcano plots of significance (Welch’s t test Benjamini and Hochberg adjusted [BH adj.] P value) versus log2(fold change) show metabolic changes in pollen-conditioned medium at time point 16 h relative to 0 h. Ions identified as pollen derived are highlighted in black. Ions annotated as glycosylated flavonoids, flavonoid aglycones (non-glycosylated flavonoids), or putative flavonoid breakdown products are shown in color when they displayed log2(fold changes) ≥ |1|. Other annotated ions are plotted in grey. (B) Model for the metabolism of flavonoids in the bee gut. Flavonoids are deglycosylated by specific bee gut bacteria, resulting in the release of flavonoid aglycones. The sugar residues are likely fermented into organic acids. Accumulation of several intermediates of aromatic compound degradation pathways, both in vivo and in vitro, suggests that the aglycone may be broken down further. The numerical values of the line graphs and the volcano plots can be found in S1 Data and S9 Data, respectively.

Fig 6. Cross-feeding between G. apicola and S. alvi.

(A) Evidence for cross-feeding of pyruvate in the honey bee gut. Z-score transformed ion intensities revealed that the ion annotated as pyruvate accumulated in bees mono-colonized with G. apicola but was depleted in hive bees, CL bees, and bees mono-colonized with S. alvi and Firm-5. (B) Growth improvement of S. alvi in G. apicola-conditioned medium. S. alvi was grown in pollen-conditioned medium in the presence (black line) or absence (dashed line) of G. apicola culture supernatant (50%, v/v). Growth was determined based on OD₆₀₀ at time points 0 h, 16 h, 36 h, and 72 h. n.s., not significant; * P < 0.05 (Welch’s t test, Benjamini and Hochberg adjusted [BH adj.]). (C) Six potentially cross-fed ions that accumulated during in vitro growth of G. apicola (left subpanel) and were consumed by S. alvi when it was grown in the presence G. apicola culture supernatant (right subpanel). Data from panels B and C come from the same experiment. Smoothed lines are added for interpretation purposes only in panel C and are dashed in the left subpanel because they are drawn through two points only. Error bars represent the standard deviation based on three replicate cultures. Chemical structures of the first annotation of each ion are shown. Asterisks indicate ions with ambiguous annotations. The numerical data of panel A can be retrieved from S8 Data. All other values are available in S1 Data.

Fig 7. Summary of the metabolic activities of the bee gut microbiota identified in this study.

(A) Schematic representation of the bee gut depicting the crop, midgut, and hindgut. The hindgut is divided into the ileum and the rectum, where the highest bacterial densities are found. Bacteria in the ileum are shown in magenta and orange (mostly Proteobacteria), and those in the rectum are shown in green and blue (mostly Lactobacilli and Bifidobacteria). Pollen grains are shown in yellow. (B) Pollen is likely predigested in the midgut, where bacterial levels are relatively low [45]. Here, the host absorbs accessible pollen-derived compounds such as simple sugars (glucose or fructose) and amino acids [46,47]. Nondigested pollen compounds enter the hindgut, where bacterial density is higher. We found nucleosides, various carboxylic acids (e.g., citrate, malate, and fumarate), and aromatic compounds (such as quinate) from pollen to be utilized by bee gut bacteria. In the posterior part of the hindgut (rectum), three community members (Firm-5, Firm-4, and B. asteroides) metabolize major components of the outer pollen wall, including flavonoids, phenolamides, and ω-hydroxy acids. The metabolic activities of the microbiota lead to the accumulation of fermentation products and intermediates of aromatic compound degradation. Some of the bacterial products may be utilized by other community members, as exemplified by the cross-feeding between G. apicola and S. alvi, or absorbed by the host. In addition, the gut symbiont B. asteroides seems to increase the production of several host metabolites (juvenile hormone derivatives and prostaglandins) that have key functions in immunity and physiology.