Comparative Genomics and Metabolomics in the Genus Nocardia

Traditional methods for natural product discovery relied exclusively on cultivation of bacteria in the laboratory. Discovery of novel chemistry was laborious and time consuming and frequently resulted in the rediscovery of known natural products. The increasing availability of genomic sequences represents a huge treasure trove for researchers. Biosynthetic pathways can be readily identified from sequence data using genome mining tools such as antiSMASH. However, using automated bioinformatic genome analysis tools, it is often unclear to what degree genetic variability in homologous biosynthetic pathways relates to chemical structural diversity. This makes prioritization of microbial strains and compound identification extremely difficult and often leads to incorrect prediction of natural product structural diversity. In the scope of this thesis, the metabolic potential of Nocardia, an under-investigated but highly prolific actinobacterial genus, was assessed using a genomics-guided approach. Sequence similarity networks generated by BiG-SCAPE (Biosynthetic Genes Similarity Clustering and Prospecting Engine) showed the presence of distinct gene cluster families including a plethora of biosynthetic gene clusters of various classes including: polyketide; non-ribosomal peptide; and terpenoid pathways. Highly conserved biosynthetic pathways encoding for nocobactin-type siderophores were used to exemplify how specific differences in highly related gene clusters correlate to structural diversity in the produced compounds. Metabolic profiling of selected Nocardia strains using LC-MS (liquid chromatography-mass spectrometry) metabolomics data and GNPS (Global Natural Product Social Molecular Networking) revealed related nocobactin-type biosynthetic gene clusters that can indeed be assigned to distinct structural types of nocobactin-type siderophores. The new nocobactin-type siderophores terpenibactins A-C were characterized using LC-MS, NMR (nuclear magnetic resonance) spectroscopy and bioassays. Furthermore, novel soluble forms of nocobactin-type siderophores were identified using LC-MS. Interestingly, some Nocardia strains produced mycobactin-type siderophores, where NMR spectroscopy revealed a chemical structure highly similar to known virulence factors from mycobacteria. The subsequent comparative genomics and metabolomics approach highlights the potential of the highly promising genus Nocardia and points out the constitutive role of nocobactins. It can set the foundation for future Nocardia genome mining approaches and thorough assessment of other rare actinobacteria, circumventing rediscovery of natural products and facilitating strain prioritization