The honey bee gut microbiota influences bee health and has become an important model to study the ecology and evolution of microbiota–host interactions. Yet, little is known about the phage community associated with the bee gut, despite its potential to modulate bacterial diversity or to govern important symbiotic functions. Here we analyzed two metagenomes derived from virus-like particles, analyzed the prevalence of the identified phages across 73 bacterial metagenomes from individual bees, and tested the host range of isolated phages. Our results show that the honey bee gut virome is composed of at least 118 distinct clusters corresponding to both temperate and lytic phages and representing novel genera with a large repertoire of unknown gene functions. We find that the phage community is prevalent in honey bees across space and time and targets the core members of the bee gut microbiota. The large number and high genetic diversity of the viral clusters seems to mirror the high extent of strain-level diversity in the bee gut microbiota. We isolated eight lytic phages that target the core microbiota member Bifidobacterium asteroides, but that exhibited different host ranges at the strain level, resulting in a nested interaction network of coexisting phages and bacterial strains. Collectively, our results show that the honey bee gut virome consists of a complex and diverse phage community that likely plays an important role in regulating strain-level diversity in the bee gut and that holds promise as an experimental model to study bacteria–phage dynamics in natural microbial communities.

Understanding how antimicrobial resistance spreads is critical for optimal application of new treatments. In the naturally competent human pathogen Streptococcus pneumoniae, resistance to β-lactam antibiotics is mediated by recombination events in genes encoding the target proteins, resulting in reduced drug binding affinity. However, for the front-line antibiotic amoxicillin, the exact mechanism of resistance still needs to be elucidated. Through successive rounds of transformation with genomic DNA from a clinically resistant isolate, we followed amoxicillin resistance development. Using whole genome sequencing, we showed that multiple recombination events occurred at different loci during one round of transformation. We found examples of non-contiguous recombination, and demonstrated that this could occur either through multiple D-loop formation from one donor DNA molecule, or by the integration of multiple DNA fragments. We also show that the final minimum inhibitory concentration (MIC) differs depending on recipient genome, explained by differences in the extent of recombination at key loci. Finally, through back transformations of mutant alleles and fluorescently labelled penicillin (bocillin-FL) binding assays, we confirm that pbp1a, pbp2b, pbp2x, and murM are the main resistance determinants for amoxicillin resistance, and that the order of allele uptake is important for successful resistance evolution. We conclude that recombination events are complex, and that this complexity contributes to the highly diverse genotypes of amoxicillin-resistant pneumococcal isolates.

Understanding how antimicrobial resistance spreads is critical for optimal application of new treatments. In the naturally competent human pathogen Streptococcus pneumoniae, resistance to β-lactam antibiotics is mediated by recombination events in genes encoding the target proteins, resulting in reduced drug binding affinity. However, for the front-line antibiotic amoxicillin, the exact mechanism of resistance still needs to be elucidated. Through successive rounds of transformation with genomic DNA from a clinically resistant isolate, we followed amoxicillin resistance development. Using whole genome sequencing, we showed that multiple recombination events occurred at different loci during one round of transformation. We found examples of non-contiguous recombination, and demonstrated for the first time that this can occur through multiple D-loop formation from one donor DNA molecule or by the uptake of multiple DNA fragments. We also show that the final minimum inhibitory concentration differs depending on recipient genome, and the sampled fitness landscape. Finally, through back transformations of mutant alleles and fluorescently labelled penicillin (bocillin-FL) binding assays, we show that pbp1a, pbp2b, pbp2x, and murM are the main resistance determinants for amoxicillin resistance, and that the order of allele uptake matters. We conclude that recombination events are complex, and that this complexity contributes to the highly diverse genotypes of amoxicillin-resistant pneumococcal isolates.