Identification of the Biosynthetic Gene Cluster of Thermoactinoamides and Discovery of New Congeners by Integrated Genome Mining and MS-Based Molecular Networking

Abstract

The putative non-ribosomal peptide synthetase (NRPS) gene cluster encoding the biosynthesis of the bioactive cyclohexapeptide thermoactinoamide A (1) was identified in Thermoactinomyces vulgaris DSM 43016. Based on an in silico prediction, the biosynthetic operon was shown to contain two trimodular NRPSs, designated as ThdA and ThdB, respectively. Chemical analysis of a bacterial crude extract showed the presence of thermoactinoamide A (1), thereby supporting this biosynthetic hypothesis. Notably, integrating genome mining with a LC-HRMS/MS molecular networking-based investigation of the microbial metabolome, we succeeded in the identification of 10 structural variants (2–11) of thermoactinoamide A (1), five of them being new compounds (thermoactinoamides G-K, 7–11). As only one thermoactinoamide operon was found in T. vulgaris, it can be assumed that all thermoactinoamide congeners are assembled by the same multimodular NRPS system. In light of these findings, we suggest that the thermoactinoamide synthetase is able to create chemical diversity, combining the relaxed substrate selectivity of some adenylation domains with the iterative and/or alternative use of specific modules. In the frame of our screening program to discover antitumor natural products, thermoactinoamide A (1) was shown to exert a moderate growth-inhibitory effect in BxPC-3 cancer cells in the low micromolar range, while being inactive in PANC-1 and 3AB-OS solid tumor models.