Evolution of an endofungal Lifestyle: Deductions from the Burkholderia rhizoxinica Genome

BACKGROUND: Burkholderia rhizoxinica is an intracellular symbiont of the phytopathogenic zygomycete Rhizopus microsporus, the causative agent of rice seedling blight. The endosymbiont produces the antimitotic macrolide rhizoxin for its host. It is vertically transmitted within vegetative spores and is essential for spore formation of the fungus. To shed light on the evolution and genetic potential of this model organism, we analysed the whole genome of B. rhizoxinica HKI 0454 - a type strain of endofungal Burkholderia species. RESULTS: The genome consists of a structurally conserved chromosome and two plasmids. Compared to free-living Burkholderia species, the genome is smaller in size and harbors less transcriptional regulator genes. Instead, we observed accumulation of transposons over the genome. Prediction of primary metabolic pathways and transporters suggests that endosymbionts consume host metabolites like citrate, but might deliver some amino acids and cofactors to the host. The rhizoxin biosynthesis gene cluster shows evolutionary traces of horizontal gene transfer. Furthermore, we analysed gene clusters coding for nonribosomal peptide synthetases (NRPS). Notably, B. rhizoxinica lacks common genes which are dedicated to quorum sensing systems, but is equipped with a large number of virulence-related factors and putative type III effectors. CONCLUSIONS: B. rhizoxinica is the first endofungal bacterium, whose genome has been sequenced. Here, we present models of evolution, metabolism and tools for host-symbiont interaction of the endofungal bacterium deduced from whole genome analyses. Genome size and structure suggest that B. rhizoxinica is in an early phase of adaptation to the intracellular lifestyle (genome in transition). By analysis of tranporters and metabolic pathways we predict how metabolites might be exchanged between the symbiont and its host. Gene clusters for biosynthesis of secondary metabolites represent novel targets for genomic mining of cryptic natural products. In silico analyses of virulence-associated genes, secreted proteins and effectors might inspire future studies on molecular mechanisms underlying bacterial-fungal interaction.