Routes of Acquisition of the Gut Microbiota of the Honey Bee Apis mellifera

Abstract

Studies of newly emerged Apis mellifera worker bees have demonstrated that their guts are colonized by a consistent core microbiota within several days of eclosure. We conducted experiments aimed at illuminating the transmission routes and spatiotemporal colonization dynamics of this microbiota. Experimental groups of newly emerged workers were maintained in cup cages and exposed to different potential transmission sources. Colonization patterns were evaluated using quantitative real-time PCR (qPCR) to assess community sizes and using deep sequencing of 16S rRNA gene amplicons to assess community composition. In addition, we monitored the establishment of the ileum and rectum communities within workers sampled over time from natural hive conditions. The study verified that workers initially lack gut bacteria and gain large characteristic communities in the ileum and rectum within 4 to 6 days within hives. Typical communities, resembling those of workers within hives, were established in the presence of nurse workers or nurse worker fecal material, and atypical communities of noncore or highly skewed compositions were established when workers were exposed only to oral trophallaxis or hive components (comb, honey, bee bread). The core species of Gram-negative bacteria, Snodgrassella alvi, Gilliamella apicola, and Frischella perrara, were dependent on the presence of nurses or hindgut material, whereas some Gram-positive species were more often transferred through exposure to hive components. These results indicate aspects of the colony life cycle and behavior that are key to the propagation of the characteristic honey bee gut microbiota.

FIG 1

Bacterial communities established through different transmission route treatments.(a and b) Individual (a) or mean (b) community composition based on 72 high-abundance OTUs binned to species (phylotype) of representative sequence. (c) Heat map of the frequency of species (with ≥1% relative abundance) within treatment groups. (d) Total 16S rRNA gene copies for treatment groups. Box plots are of quantiles, and diamonds signify mean copy number. “ϴ,” unexposed NEWs. Letters to the right of the plot demonstrate shared significance groups (Mann-Whitney).

FIG 2

Colonization dynamics within two sampled hives (upper and lower halves of figure) for ileum (left) and rectum (right). A time course is shown for relative abundance (top), total 16S gene copies (middle), and transformed absolute abundance (bottom). The transformed absolute abundances are total 16S rRNA gene copies transformed by relative abundance percentages, which provides a crude metric showing how numbers of each species change with worker age. Atypic., atypical; Env_Firm*, environmentally associated Firmicutes; Lacto, Lactobacillus.

FIG 3

Alpha diversity measures of transmission route treatment groups. (a and b) Average richness (“Shannon_Index”) (a) or evenness (“Equitability”) (b) of treatment groups. Box plots are of quantiles, and diamonds signify mean values. Letters above the plot demonstrate shared significance groups (Tukey's HSD).

FIG 4

(a) Distances among communities of bees with different exposure treatments, represented as PCoA plots of jackknifed weighted UniFrac distances. The blue ellipse shows the area of plot dominated by the NH and F treatment groups. The green ellipse shows the area dominated by the HG and to a lesser extent N and NF treatments. (b) PCoA plot of jackknifed weighted UniFrac distances of t = 8-day samples in within-hive colonization experiment (sampled at an even depth of 3,500 sequences). Samples from the ileum and rectum are fully segregated and differ significantly at 8 days posteclosure (Adonis PERMANOVA, R² = 0.42; F_1,18 = 13.29; P < 0.001). (c) Simplified PCoA plots from within-hive colonization dynamics experiment. A detailed version of this plot with all points sampled is available in figure SI3b in the supplemental material.

FIG 5

Average weighted UniFrac distances between representative control sample (AZ125.4) and transmission route treatment groups. Letters above the chart demonstrate shared significance groups at P < 0.05 evaluated with a 2-sided nonparametric t test with 1,000 Monte Carlo iterations, Bonferroni corrected.