Functional diversity within the simple gut microbiota of the honey bee

Abstract

Animals living in social communities typically harbor a characteristic gut microbiota important for nutrition and pathogen defense. Accordingly, in the gut of the honey bee, Apis mellifera, a distinctive microbial community, composed of a taxonomically restricted set of species specific to social bees, has been identified. Despite the ecological and economical importance of honey bees and the increasing concern about population declines, the role of their gut symbionts for colony health and nutrition is unknown. Here, we sequenced the metagenome of the gut microbiota of honey bees. Unexpectedly, we found a remarkable degree of genetic diversity within the few bacterial species colonizing the bee gut. Comparative analysis of gene contents suggests that different species harbor distinct functional capabilities linked to host interaction, biofilm formation, and carbohydrate breakdown. Whereas the former two functions could be critical for pathogen defense and immunity, the latter one might assist nutrient utilization. In a γ-proteobacterial species, we identified genes encoding pectin-degrading enzymes likely involved in the breakdown of pollen walls. Experimental investigation showed that this activity is restricted to a subset of strains of this species providing evidence for niche specialization. Long-standing association of these gut symbionts with their hosts, favored by the eusocial lifestyle of honey bees, might have promoted the genetic and functional diversification of these bee-specific bacteria. Besides revealing insights into mutualistic functions governed by the microbiota of this important pollinator, our findings indicate that the honey bee can serve as a model for understanding more complex gut-associated microbial communities.

Fig. 1.

Identification of bacterial species or species groups in the honey bee microbiome and analysis of their genetic diversity. Phylogenetic profile based on (A) classification of 576,192 reads mapping against 31 marker proteins with MetaPhyler (48) and (B) best BLASTP hit distribution of all 112,128 CDSs. n.a., reads or CDSs not assigned. (C) Maximum-likelihood protein phylogeny of UvrC. All eight phylogenies (see SI Appendix, Fig. S1 for other trees) revealed that most sequences from the honey bee microbiome (shown in pink) fall into the same six distinct clusters. These phylogenetic clusters are referred to as Alpha-1, Alpha-2, Snodgrassella, Gamma, Bifido, and Firm. We considered all closely related taxa with available genomes for this analysis. Bootstrap values >80 are shown. (D) Percentages of the minimal gene set present in each bin are depicted in parentheses (only full-length copies/including fragmented genes). Graphs show distribution of genes of the minimal gene set based on identified full-length copies per bin. Asterisks indicate fragmented genes. (E) Average percentage of variable sites and average read coverage for 27 ribosomal protein-encoding genes of each bin.

Fig. 2.

Functions enriched in the honey bee microbiome relative to nine other gut microbiomes. These represent the 20 most enriched functions from a total of 72 COG functions that received significant enrichment scores in eight of nine comparisons (SI Appendix, Table S2). Average D-score represents the mean enrichment score over all nine comparisons. For each function, the total number of genes found in the metagenomic dataset and the distribution in the six bins is shown. Letters above graphs indicate COG category. “G” and “K” stand for “carbohydrate metabolism and transport” and “transcription,” respectively.

Fig. 3.

Pectin degradation by the honey bee gut microbiota. (A) Proteins involved in pectin catabolism, for which respective genes were identified, are indicated in red. The illustrated pectin molecule shows the homo- and rhamnogalacturonate backbone with side chains (50) and cleavage sites for identified GHs and PLs. PaeY indicates a pectin acetyl esterase removing ester residues. Except for proteins marked with an asterisk, the encoding genes were present in the Gamma bin. Genes encoding the oligogalacturonide-specific porin KdgM, the inner-membrane transporters TogT, ExuT, and KdgT, and a pathway for galacturonide conversion were also identified. (B) Four different loci assigned to the Gamma bin encoding key proteins for pectin breakdown (scaffold numbers are indicated). The first two loci encode proteins involved in metabolism of PGA. The latter two might be involved in breakdown and metabolism of xylose side chains. Blue, green, magenta, and yellow indicate functions associated with glycoside cleavage, transport, cellular catabolism, and regulation, respectively. (C) PGA degradation is detected by the formation of clearance zones around bacterial lawns. The species to which each bacterial isolate belongs is indicated. γ, Gilliamella; β, Snodgrassella; α1, Alpha-1; and B, Bifido. (D) Maximum-likelihood protein phylogeny of metagenomic pectate lyases of the PL1 family and their closest homologs found in other bacteria as well as outgroups Aspergillus niger and Arabidopsis thaliana. Bootstrap values >80 and ecological niche of bacteria are indicated.