Seed-applied multi-kingdom synthetic communities selectively reshape bacterial communities and highlight key criteria for strain selection
Defined microbial communities (also known as synthetic communities) are showing promising results for plant health but are lacking an efficient lab to field transition. This is the due to limited knowledge on how to efficiently modulate plant microbiota by considering complex environments and multi-kingdom interactions. In this study, we aimed to better understand the transmission and impact of multi-kingdom synthetic communities (SynComs) from the seed to seedling stage. We constructed 20 different SynComs using both a priori and random approaches, from pool of diverse strains including 24 bacteria, 11 yeasts and 10 filamentous fungi. SynComs were inoculated on Brassica napus seeds and we monitored both transmission and impact on the microbiota of 15-day-old seedlings grown in non-sterile soil. Optimization of the inoculation protocol showed that alginate coating improved bacterial, yeast and filamentous fungi concentrations by more than 2 log compared with other approaches. With this inoculation method, we observed contrasted seedling colonization profiles, with SynComs members representing between 1.1-45.7% of bacterial community and 3.2-36.6% of fungal community. Our multiple SynCom design revealed that strain selection is a more critical determinant of SynCom performance than assembly strategy. Even randomly assembled communities performed well, as long as they are drawn from a pool of ecologically relevant, well-adapted taxa. Based on these evidences, we identified key bacterial and fungal traits explaining efficient seedling colonization such as high abundance on inoculated seed and low in vitro lag-time. Despite low colonization levels, we observed that SynCom inoculation altered seeding bacterial community assembly in 14 SynComs. A total of 82 native bacterial ASVs were identified as responsive to SynCom inoculation, most likely originating from the soil. This shift indicates that SynComs influence community assembly by modulating the recruitment of environmental taxa, especially when SynCom strains were more integrated in multi-kingdom network structures. Finally, we identified four distinct SynComs profiles which either colonized strongly or not seedlings while shifting of not native microbiota. Altogether, these findings provide actionable directions for improving SynCom design, suggesting that leveraging ecological processes such as host adaptation, optimal inoculation density, and network integration could enhance both colonization efficiency and plant phenotypic outcomes.