Mapping phyllosphere microbiota interactions in planta to establish genotype-phenotype relationships

Bacteria-bacteria interactions and resultant microbiota changes were characterized using a synthetic phyllosphere community in planta. Host-associated microbiomes harbour hundreds of bacterial species that co-occur, creating the opportunity for manifold bacteria-bacteria interactions, which in turn contribute to the overall community structure. The mechanisms that underlie this self-organization among bacteria remain largely elusive. Here, we studied bacterial interactions in the phyllosphere microbiota. We screened for microbial interactions in planta by adding 200 endogenous strains individually to a 15-member synthetic community and tracking changes in community composition upon colonization of the model plant Arabidopsis. Ninety percent of the identified interactions in planta were negative, and phylogenetically closely related strains elicited consistent effects on the synthetic community, providing support for trait conservation. Community changes could be largely explained by binary interactions; however, we also identified a higher-order interaction of more than two interacting strains. We further focused on a prominent interaction between two members of the Actinobacteria. In the presence of Aeromicrobium Leaf245, the population of Nocardioides Leaf374 was reduced by almost two orders of magnitude. We identified a potent antimicrobial peptidase in Aeromicrobium Leaf245, which resulted in Nocardioides Leaf374 lysis. A respective Leaf245 mutant strain was necessary and sufficient to restore Nocardioides colonization in planta, demonstrating that direct bacteria-bacteria interactions were responsible for the population shift originally observed. Our study highlights the power of synthetic community screening and uncovers a strong microbial interaction that occurs despite a spatially heterogeneous environment.

Automated summary

This study investigates bacterial interactions in the phyllosphere microbiota using a synthetic community. The results show that 90% of the interactions in planta were negative and closely related strains had consistent effects on the synthetic community. The changes in the community could be largely explained by binary interactions, but higher-order interactions involving more than two strains were also identified.