Glycopeptide antibiotics (GPAs) such as teicoplanin and vancomycin are of great pharmacological relevance. These bioactive molecules are produced in vivo by a multi-enzyme complex known as a non-ribosomal peptide synthetase (NRPS)(1)(2)(3)(4). The typical linear NRPS consists of repeating modules that can be further divided into different catalytically active domains, with each module incorporating one amino acid into the growing peptide chain. The key domains present in typical NRPS modules are adenylation (A)-domains, peptidyl-carrier (PCP)-domains, condensation (C)-domains and a thioesterase (TE)-domain, of which the latter is only present within terminating NRPS modules(5)(6). After the NRPS-catalysed formation of the linear peptide intermediate GPAs undergo further maturation processes, including the introduction of oxidative crosslinks through the action of tailoring enzymes (P450 monooxygenases) to obtain their final biological activity(4). Within NRPS biosynthesis, there is still much debate over the exact origins of the selectivity of NRPS machineries – it is essential to answer these questions if we are to ever realise the goal of reengineering these systems to produce novel compounds.
Due to the importance of the final stages in peptide biosynthesis, we have concentrated on characterising the activity of the final module of the teicoplanin NRPS (Tcp12). We have developed assays to assess the activity of both A-domains (via a novel spectroscopic coupled enzyme assay)(7) and C-domains, where we have applied Tcp12. Our results reveal that the selectivity of peptide bond formation during NRPS biosynthesis is largely enforced by monomer selection mediated by A-domains, with C-domains displaying much lower selectivity. These findings greatly improve the prospects for producing new GPAs by NRPS redesign centred on mutations to A-domains.