Discovery, Characterisation and Engineering of Non-Ribosomal Peptide Synthetases and Phosphopantetheinyl Transferase Enzymes
Non-ribosomal peptide synthetases (NRPSs) are multi-modular biosynthetic enzymes that are responsible for the production of many bioactive secondary metabolites produced by microorganisms. They are activated by phosphopantetheinyl transferase (PPTase) enzymes, which attach an essential prosthetic group to a specific site within a “carrier protein” (CP) domain that is an integral part of each NRPS module. Of particular importance in this work is the NRPS BpsA, which produces a blue pigment called indigoidine; but only when BpsA has first been activated by a PPTase. BpsA can be used as a reporter for PPTase activity, to identify PPTases and/or measure their activity. Several CP-substituted BpsA variants were used, in order to study and identify PPTases which may recognise different CP domains. The first part of the research described in this thesis examined the features of foreign CP interactions within BpsA that made these functional substitutions possible. Two key residues, the +4 and +24 positions relative to an invariant serine, were found to be highly important; with appropriate substitutions at these positions yielding active CP-substituted variants. Wild type BpsA and the CP-substituted variants were then used as the basis of a screen to discover new PPTase genes, and associated natural product biosynthetic genes, from metagenomic libraries. The vast majority of bacteria that produce bioactive secondary metabolites are unable to be cultured under laboratory conditions; screening metagenomic libraries is a way to access this untapped biodiversity in order to discover new natural products. Two environmental DNA libraries were screened, and PPTase genes were identified via their ability to activate BpsA, giving rise to blue colonies in high throughput agar plate screens. This screen proved to be a powerful enrichment strategy with almost half of the novel 21 PPTase genes recovered also linked to biosynthetic gene clusters. Using the evolved CP-substituted BpsA variants (and thereby altering the PPTase recognition site) enabled a wider variety of hits to be found. This led to the hypothesis that some of the PPTases discovered via this screening method would have non-overlapping substrate specificities, a beneficial property for certain PPTase applications. The 21 PPTase genes discovered via metagenomic screening were characterised further, using a series of assays involving BpsA to measure their activity. As is common for PPTase enzymes, there were difficulties in obtaining enough soluble protein via purification to perform a detailed analysis of each. Those that were able to be purified had much lower activity than other previously characterised PPTases, and were also not as specific for their CP substrates as they had first appeared to be. Due to these low activity levels, several other previously characterised PPTases were also studied further using the BpsA methods. All PPTases showed a relatively broad activity across a range of CP substrates. The desire to obtain PPTases with more specific substrate specificities led to the development of a directed evolution screen to alter PPTase CP specificity. In a proof-of-principle study the E. coli PPTase EntD was evolved to lose activity with the BpsA CP while retaining activity with its native CP. This screen, the first of its kind to evolve PPTases for greater CP substrate specificity, was successful in recovering several improved variants. These variants had either completely abolished or vastly decreased activity for the WT BpsA CP while retaining the ability to activate the native (EntF) CP domain. The general strategy developed here can be applied to the evolution of other PPTases and CP substrates.