Interspecies interactions can drive the emergence of unexpected microbial phenotypes that are not observed when studying monocultures. The cystic fibrosis (CF) lung consists of a complex environment where microbes, living as polymicrobial biofilm-like communities, are associated with negative clinical outcomes for persons with CF (pwCF). However, the current lack of in vitro models integrating the microbial diversity observed in the CF airway hampers our understanding of why polymicrobial communities are recalcitrant to therapy in this disease. Here, integrating computational approaches informed by clinical data, we built a mixed community of clinical relevance to the CF lung composed of Pseudomonas aeruginosa, Staphylococcus aureus, Streptococcus sanguinis, and Prevotella melaninogenica. We developed and validated this model biofilm community with multiple isolates of these four genera. When challenged with tobramycin, a front-line antimicrobial used to treat pwCF, the microorganisms in the polymicrobial community show altered sensitivity to this antibiotic compared to monospecies biofilms. We observed that wild-type P. aeruginosa is sensitized to tobramycin in a mixed community versus monoculture, and this observation holds across a range of community relative abundances. We also report that LasR loss-of-function, a variant frequently detected in the CF airway, drives tolerance of P. aeruginosa to tobramycin specifically in the mixed community. Our data suggest that the molecular basis of this community-specific recalcitrance to tobramycin for the P. aeruginosa lasR mutant is increased production of phenazines. Our work supports the importance of studying a clinically relevant model of polymicrobial biofilms to understand community-specific traits relevant to infections.

While foraging for nectar and pollen, bees are exposed to a myriad of xenobiotics, including plant metabolites, which may exert a wide range of effects on their health. Although the bee genome encodes enzymes that help in the metabolism of xenobiotics, it has lower detoxification gene diversity than the genomes of other insects. Therefore, bees may rely on other components that shape their physiology, such as the microbiota, to degrade potentially toxic molecules. In this study, we show that amygdalin, a cyanogenic glycoside found in honey bee-pollinated almond trees, can be metabolized by both bees and members of the gut microbiota. In microbiota-deprived bees, amygdalin is degraded into prunasin, leading to prunasin accumulation in the midgut and hindgut. In microbiota-colonized bees, on the other hand, amygdalin is degraded even further, and prunasin does not accumulate in the gut, suggesting that the microbiota contribute to the full degradation of amygdalin into hydrogen cyanide. In vitro experiments demonstrated that amygdalin degradation by bee gut bacteria is strain-specific and not characteristic of a particular genus or species. We found strains of Bifidobacterium, Bombilactobacillus, and Gilliamella that can degrade amygdalin. The degradation mechanism appears to vary since only some strains produce prunasin as an intermediate. Finally, we investigated the basis of degradation in Bifidobacterium wkB204, a strain that fully degrades amygdalin. We found overexpression and secretion of several carbohydrate-degrading enzymes, including one in glycoside hydrolase family 3 (GH3). We expressed this GH3 in Escherichia coli and detected prunasin as a byproduct when cell lysates were cultured with amygdalin, supporting its contribution to amygdalin degradation. These findings demonstrate that both host and microbiota can act together to metabolize dietary plant metabolites.