Chemoenzymatic Synthesis of Chromodepsipeptides and Natural Product Discovery via Genome Mining
Recent advances in the development of sequencing technologies have enabled the identification of a multitude of bacterial gene clusters, putatively involved in the biosynthesis of nonribosomal peptides (NRPs). Peptides of nonribosomal origin constitute a class of structurally and functionally divers...
|PDF Full Text
No Tags, Be the first to tag this record!
|Recent advances in the development of sequencing technologies have enabled the identification of a multitude of bacterial gene clusters, putatively involved in the biosynthesis of nonribosomal peptides (NRPs). Peptides of nonribosomal origin constitute a class of structurally and functionally diverse natural products, which are assembled by multimodular nonribosomal peptide synthetases (NRPSs). These compounds exhibit a broad pharmacological spectrum, ranging from antibacterial- to immunosuppressive properties. Understanding the assembly mechanisms in combination with rational genome mining approaches will provide opportunities for the discovery of new bioactive natural products.
Within this study one approach was utilized to generate thiocoraline analogs via chemoenzymatic synthesis and the second strategy focused on the de novo natural product discovery via genome mining.
Thiocoraline represents a pseudosymmetrical chromophore-capped octathiodepsipeptide, in which the symmetrical halves are linked via thioester bonds. In this study, the cyclodimerization potential of the thioesterase domain of the thiocoraline biosynthetic machinery (TioS PCP-TE) was investigated to obtain further insights into the iterative assembly of chromodepsipeptides. To address this objective, the recombinant enzyme was incubated with synthetically derived tetrapeptidyl substrates, resembling thiocoraline precursors. It was shown that the enzyme catalyzes the cyclodimerization of linear precursor molecules and an unprecedented macrothiolactonization. Evaluation of the biocombinatorial potential established the thioesterase as a robust and versatile catalyst for the generation of chromodepsipeptide analogs, harbouring thioester- or ester-linkages. As thiocoraline attains its antitumor activity from DNA-bisintercalation, the chemoenzymatically generated macrocycles were isolated and investigated towards DNA-bisintercalation activity in vitro.
In the second part of this study, bioinformatic analysis of the 8.2 Mb Saccharopolyspora erythraea genome revealed two cryptic NRPS gene clusters related to hydroxamate-type siderophore biosynthesis. Detailed analysis of adenylation domain substrate-specificity and module organization enabled the establishment of a highly selective and sensitive radio-LCMS-guided genome mining approach. Application of this approach resulted in the discovery of the siderophore erythrochelin. Structure elucidation of erythrochelin was accomplished via NMR- and MSn-analysis and revealed the sequence of the tetrapeptide siderophore to be: α-N-acetyl-δ-N-acetyl-δ-N-hydroxy-D-ornithine-D-serine-cyclo(δ-N-hydroxy-L-ornithine-δ-N-acetyl-δ-N-hydroxy-L-ornithine).
Erythrochelin assembly requires the proliferation of δ-N-hydroxy-L-ornithine (L-hOrn) and δ-N-acetyl-δ-N-hydroxy-L-ornithine (L-haOrn). The corresponding modifying enzymes, the FAD-dependent monooxygenases EtcB and Sace_1309 together with the bifunctional malonyl-CoA decarboxylase/N-acetyltransferase were identified and biochemically characterized. In vitro studies revealed EtcB and Sace_1309 to exclusively catalyze the δ-N-hydroxylation of free L-ornithine. The second tailoring enzyme, Mcd, was shown to catalyze malonyl-CoA decarboxylation and subsequent acetyltransfer onto the δ-hydroxamino group of L-hOrn, affording L-haOrn. Based on the elucidation of precursor biosynthesis (L-haOrn), a model for the entire erythrochelin assembly is presented.