Natural Products DOI: 10.1002/anie.201106305 Myxoprincomide: A Natural Product from Myxococcus xanthus Discovered by Comprehensive Analysis of the Secondary Metabolome** NiÇa Socorro Cortina, Daniel Krug, Alberto Plaza, Ole Revermann, and Rolf Müller* Bacteria are important sources of therapeutically relevant natural products. [1] Early established bacterial producers include the streptomycetes, pseudomonads, and bacilli, but the list has recently expanded to include further sources, such as the myxobacteria. [2–5] Access to whole genome sequences has indicated that among bacteria producing natural products through the action of polyketide synthase (PKS) and non- ribosomal peptide synthetase (NRPS) the number of known compounds is remarkably lower than the genetic capacity of the organism for secondary-metabolite biosynthesis. [1, 6] Thus, identification of new compound classes and their assignment to biosynthetic gene clusters is a crucial step in the discovery of novel natural products. Here, we report the discovery and structural elucidation of myxoprincomide (1), a novel NRPS/PKS natural product from Myxococcus xanthus DK1622, by combining methods of targeted mutagenesis, liquid chromatography coupled to high-resolution mass spectrometry (LC-HRMS), and statis- tical data evaluation (Figure 1). Myxoprincomide (1), a linear peptide bearing some unusual residues, is produced by the NRPS/PKS biosynthetic machinery encoded by the mxp (MXAN_3779) gene locus. Moreover, by correlating addi- tional very-low-abundance natural products to biosynthetic pathways in DK1622 we demonstrate how a comprehensive “metabolome-mining” approach can complement genome- mining strategies in the discovery of secondary metabolites. The myxobacterial strain DK1622, considered a model organism for the study of myxobacterial social motility and multicellular differentiation, was not recognized as a pro- ducer of secondary metabolites until its genome was sequenced. [7] However, interest in resolving its secondary metabolome has increased since bioinformatic analysis revealed it to contain 18 biosynthetic gene clusters encoding NRPS, PKS, and NRPS/PKS hybrid systems. To date only five compound classes derived from NRPS and PKS biosynthetic machineries have been characterized and correlated to their gene clusters in DK1622 (see Figure 1 in the Supporting Information). [8–13] Intriguingly, evaluation at the transcrip- tomic and proteomic levels asserts that the remaining 13 unassigned pathways are active under standard conditions for the cultivation of DK1622. [14, 15] We reasoned that the low abundance of the corresponding compounds may have previously prevented their detection, and thus set out to find these compounds by utilizing advanced analytical methods. The available DK1622 genome sequence facilitated the construction of a targeted mutant library including knockouts of every secondary-metabolite biosynthetic gene cluster (Table 1 in the Supporting Information). [16] As the process of finding the possibly subtle differences between wild-type and mutant secondary-metabolite profiles by manual com- parison of LC-MS data is frequently tedious, error-prone, and low in sensitivity, we sought to implement statistical tools in order to expedite our LC-MS data evaluation. In preparation for the comprehensive statistical analysis, mutant and wild- type strains were grown in small-scale fermentation in quadruplicate, replicate extracts were analyzed by LC- HRMS, and data were pretreated by using a compound- finding algorithm, resulting in the definition of > 1000 molecular features per sample. [17–19] In order to identify molecular features specifically missing in culture extracts from DK1622 mutant strains, we applied principal-compo- nent analysis (PCA) to the preprocessed LC-MS datasets [*] M. Sc. N. S. Cortina, [+] Dr. D. Krug, [+] Dr. A. Plaza, Dipl.-Chem. O. Revermann, Prof. Dr. R. Müller Abteilung Mikrobielle Naturstoffe, Helmholtz-Institut für Pharmazeutische Forschung Saarland (HIPS), Helmholtz-Zentrum für Infektionsforschung and Institut für Pharmazeutische Biotechnologie, Universität des Saarlandes Campus, Gebäude C2.3, 66123 Saarbrücken (Deutschland) E-mail: rom@helmholtz-hzi.de Homepage: http://www.helmholtz-hzi.de/hips [ + ] These authors contributed equally to this work. [**] We thank Gregory Velicer of Indiana University for providing strain Myxococcus xanthus A2 and Charles Moore for critically reading this manuscript. We thank Thomas Hoffmann and Eva Luxenburger for help with the HPLC-MS analyses, Wolfgang Kessler and Stephan Hüttel for the large-scale fermentation, and Gabriela Zurek and Aiko Barsch (Bruker Daltonik) for their advice. N.S.C. thanks the Deutscher Akademischer Austauschdienst for a fellowship. Research in the R.M. group was funded by the Bundesministerium für Bildung und Forschung and the Deutsche Forschungsgemein- schaft. Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/anie.201106305. Angewandte Chemie 811 Angew. Chem. Int. Ed. 2012, 51, 811 –816 # 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim