Metarhizium is a genus of endophytic, insect-pathogenic fungi that is used as a biological control agent. The dual lifestyles of these fungi combine the parasitism of insect pests with the symbiotic association with plant roots. A major class of secreted metabolites by Metarhizium are cyclic depsipeptides called destruxins (DTXs). As prominent insecticidal compounds, their role during plant interactions is still largely unknown. Here, we examined the metabolomic profile of Metarhizium, with special emphasis on DTX production, using untargeted, liquid chromatography-tandem mass spectrometry (LC-MS/MS). Four Metarhizium species, two insect generalists (M. robertsii and M. brunneum), and two insect specialists (M. flavoviride and M. acridum) were inoculated onto agar plate cultures containing either bean (Phaseolus vulgaris) or corn (Zea mays) and grown for four and seven days. After methanol extraction, feature-based molecular networking (FBMN) was used to obtain DTX identification as defined by the Global Natural Products Social Molecular Networking (GNPS). A total of 25 DTX analogs were identified, with several DTX-like compounds in coculture that could not be identified. Metarhizium species differed in the amount and type of DTXs they produced, with the insect specialists producing far fewer amounts and types of DTXs than the insect generalists. The production of these metabolites varied between cultures of different ages and plant hosts. Conditions that influence the production of DTXs are discussed. As the genetic arsenal of natural products relates to the lifestyle of the organism, uncovering conditions with an ecological context may reveal strategies for producing novel compounds or precursors suitable for synthetic biology. IMPORTANCE The development of an intimate and beneficial association between fungi and plants requires an exchange of a complex mixture of chemical cues. These compounds are a means of communication, promoting or limiting the interaction, but can have numerous other biological and ecological functions. Determining how the metabolome, or a subset thereof, is linked to plant host preference and colonization has implications for future functional studies and may uncover novel therapeutic compounds whose production is elicited only under cocultivation. In this study, we performed an untargeted metabolomic analysis of plate cocultures with individual plant-fungal pairs. The identification of a major group of fungal metabolites, the destruxins, was examined for their role in plant specificity. The diversity of these metabolites and the production of numerous unidentified, structural analogs are evidence of the sensitivity of the methodology and the potential for future mining of this living data set.
Keywords: GNPS; MS-DIAL; MetaboAnalyst; endophyte; feature-based molecular networking; fungal metabolites; metabolomics.