Abstract
Abstract
Plant-associated microbiomes play a crucial role in sustaining host health and fitness, but how pathogen invasion reshapes microbial interaction networks and how to leverage this for developing effective biocontrol strategies remain poorly understood. This study investigated the maize microbiome under Fusarium stalk rot stress by integrating large-scale amplicon sequencing, network refinement, and experimental validation. To distinguish biological interactions from spurious associations driven by environmental and geographic factors, we applied a refined analytical framework to filter out abiotic correlations. The resulting interaction networks revealed that pathogen infection significantly increased network complexity and the proportion of potential negative interactions across all maize compartments except seeds. Specifically, potential negative interactions with Fusarium were significantly enriched in the rhizosphere microbiome, encompassing numerous uncultured taxa and known biocontrol genera (e.g., Epicoccum, Burkholderia, and Nocardioides). Representative strains from these taxa were subsequently isolated and assembled into a synthetic microbial community (SynCom). This refined network-derived SynCom significantly suppressed maize stalk rot caused by Fusarium verticillioides and F. graminearum. In non-sterilized soil, this SynCom exhibited significantly stronger disease-suppressive effects than either the randomly assembled SynCom or the SynCom derived from the unrefined network. Integrated transcriptomic and metabolomic analyses suggested that the SynCom confers resistance through the production of antifungal metabolites, including azelaic acid, and potentially through the induction of host defense responses. Collectively, our study provides meaningful insights into plant-microbe interaction under pathogen stress and suggests an analytical–experimental workflow that integrates computational ecology with experimental validation for developing biocontrol strategies against maize stalk rot.