A combined bioinformatic and mass spectrometric approach to guide natural product discovery and production

Synthetic biology approaches have proven to be highly useful for efficient, large-scale production of known natural products; however, their use for the discovery of new compounds can be limited by the unknown nature of target molecule. Thus, without knowledge of the small molecule product, it is difficult assess the efficacy of synthetic constructs designed to express novel natural product gene clusters. Moreover, synthetic constructs often fail due to poor heterologous expression of subsets of genes within the target pathway. In these cases, it would be very helpful to have a set of related genes for all target pathways. We have developed a combined bioinformatics/mass spectrometric approach that addresses both of these issues. Accordingly, bioinformatics-based grouping of natural product gene clusters into families across hundreds of genomes allows prioritization of discovery efforts, while creating a parts list that can be applied to synthetic design strategies. Further, the correlation of specific metabolites with the presence of specific gene cluster families in large numbers of strains allows discovery of the molecules produced by individual gene clusters, thus establishing the correct target for production efforts.