Abstract
Tobacco bacterial wilt, caused by Ralstonia solanacearum, threatens global tobacco production. While the rhizosphere microbiome defends against soil-borne pathogens, mechanisms underlying how bacterial wilt reshapes microbial community structure, function, and ecological interactions remain poorly understood. Here, we employed metagenomic sequencing to investigate taxonomic and functional alterations in the rhizosphere microbiome of symptomatic (S) and asymptomatic (A) tobacco plants across two locations (Fenggang and Bozhou), establishing four groups: FA, FS, BA, and BS. Quality control of sequencing data showed no technical bias between groups (p > 0.05). Contrary to the paradigm that pathogen invasion reduced microbial diversity, alpha diversity analysis revealed higher species richness (Sobs) in symptomatic soils, whereas community evenness (Shannon and Simpson indices) remained unchanged, suggesting selective reshuffling rather than microbiome collapse. Beta-diversity analysis revealed significant compositional shifts associated with disease status (PERMANOVA, R (2) = 0.713, p = 0.001), with symptomatic communities displaying greater heterogeneity. Taxonomic profiling revealed consistent enrichment of the pathogen R. solanacearum and opportunistic bacteria (including Stenotrophomonas and Pseudomonas) in symptomatic rhizospheres, concomitant with depletion of putative beneficial taxa (Candidatus_Solibacter, Luteitalea, and Metarhizium). Functional annotation indicated a metabolic shift from homeostatic maintenance to stress adaptation and pathogenicity. Symptomatic soils exhibited significant enrichment of virulence factors, including motility and secretion system genes, microbial defense mechanism genes (COG), and antibiotic resistance genes (CARD). Additionally, increased abundance of carbohydrate-active enzymes (CAZy)-particularly glycoside hydrolases-suggested intensive nutrient acquisition from decaying tissues. Co-occurrence network analysis revealed that asymptomatic communities formed denser, competition-driven networks characterized by a higher proportion of negative correlations. Disease destabilized these networks by reducing connectivity and, crucially, rewired interactions of R. solanacearum from negative to positive associations with taxa such as Sphingobium, thereby reflecting erosion of competitive constraints and pathogen incorporation into cooperative networks. Our findings revealed that bacterial wilt drove multi-layered dysbiosis, encompassing pathogen-driven taxonomic selection, functional shifts toward stress adaptation and intensified competition, and collapse of stable antagonistic networks associated with plant health. This study provided mechanistic insights into microbiome-mediated disease progression and identified specific microbial taxa and network properties as candidate targets for ecological disease management and early diagnostic indicators.