Characterization and comprehensive analysis of the ecological interaction networks of bacterial communities in Paullinia cupana var. sorbilis by 16S rRNA gene metabarcoding.

Endophytes improve the host performance in areas of high plant endemicity. Paullinia cupana is an Amazonia plant species of economic and social importance due to the high caffeine concentration in its seeds. An interesting strategy to identify endophytic microorganisms with potential biotechnological application is to understand the factors that influence the endophytic community to rationalize the host management programs. We used the next-generation sequencing for bacterial 16S rRNA gene to examine how the P. cupana organ, genotype, and geographic location influenced its endophytic bacterial community. We obtained 1520 operational taxonomic units (OTUs) distributed in 19 phyla, 32 classes, 79 orders, 114 families and 174 genera. The P. cupana roots and leaves were specifically colonized by the bacterial genera Acidothermus and Porphyromonas, respectively, with high relative frequency. The plant organ type influenced the endophytic community's richness, diversity, OTUs composition, relative abundance of phyla and genera, and genera interaction network. However, the host plant genotype and geographic location influenced the composition and interaction among genera in the network analysis. Prevotella is a super-generalist genus in the interaction network of endophytic bacteria of P. cupana. This study revealed endophytic bacterial groups of importance to P. cupana and stressed that the host plant organ modulates the structure and interactions within this community. Our results indicated that the microbial community adapted to colonize P. cupana by adjusting to its composition and interaction network. The isolation of abundant and super-generalist bacterial genera shall help to examine their functionality to the composition and fitness of the endophytic community of P. cupana.