ABSTRACT
The common soil bacteria Serratia marcescens, Serratia proteamaculans , and Bacillus subtilis produce small molecular weight volatile compounds that are fungi-static against multiple species, including the zygomycete mold Rhizopus stolonifer (Mucoromycota) and the model filamentous mold Neurospora crassa (Ascomycota). The compounds or the bacteria can be exploited in development of biological controls to prevent establishment of fungi on food and surfaces. Here, we quantified and identified bacteria-produced volatiles using headspace sampling and gas chromatography-mass spectrometry. We found that each bacterial species in culture has a unique volatile profile consisting of dozens of compounds. Using multivariate statistical approaches, we identified compounds in common or unique to each species. Our analysis suggested that three compounds, dimethyl trisulfide, anisole, and 2-undecanone, are characteristic of the volatiles emitted by these antagonistic bacteria. We developed bioassays for testing inhibition of each compound and found dimethyl trisulfide and anisole were the most potent. This work establishes a pipeline for translating volatile profiles of cultured bacteria into high quality candidate fungistatic compounds which may be useful in combination as antifungal control products. IMPORTANCE
Bacteria may benefit by producing fungistatic volatiles that limit fungal growth providing a mechanism to exclude competitors for resources. Volatile production is potentially mediating long distance biological control and competitive in-teractions among microbes, but the specific bioactive compounds are poorly characterized. This work provides evidence that fungistatic compounds in complex blends can be identified using machine-learning and multivariate approaches. This is the first step in identifying pathways responsible for fungistatic volatile production in order to phenotype and select natural strains for biocontrol ability, or engineer bacteria with relevant pathways.