Reprogramming Versus Subdomain Substitution: An Exploration of Strategies for NRPS Engineering Using the Unique Model System BpsA

<p>Non-ribosomal peptide synthetases (NRPSs) are large enzymes that generate a plethora of important natural products, from antibiotics to immunosuppressants. These modular enzymes function like an assembly line, selecting and incorporating specific (and frequently nonproteinogenic) amino acids into a growing peptide chain. This modular structure offers promise for re-engineering NRPS units to generate new useful products, but progress has to date been limited by the complex and dynamic nature of key domains, and a failure to define generally applicable “rules” to guide engineering efforts. Early efforts to engineer NRPS enzymes relied on the substitution of entire NRPS modules or domains, but product yields were often very low. However, these studies did highlight the promise of targeting the adenylation domain, the part of each NRPS modules that is responsible for selecting each amino acid substrate. Two particularly promising strategies for NRPS engineering aim to manipulate the adenylation domain in ways that minimise steric disruption to the assembly line. The first of these, reprogramming, makes the fewest possible changes to the NRPS primary sequence, but is dependent on those precise changes conforming to the existing structure of the adenylation domain binding pocket. More recently a second technique has been developed, subdomain substitution, which recombines a larger region of the adenylation domain to avoid perturbation of the binding pocket. The research described in this thesis examined and compared both approaches using the unique NRPS BpsA as a model system. BpsA is a single-module NRPS that generates a vivid blue pigment product, making for a reductionist system that offers a robust visual reporter capacity. Experiments with the reprogramming technique showed that small changes to the protein sequence had potential to exert major impacts on enzyme function, even when no change to function was intended. In contrast, experiments with subdomain substitution were generally more effective, showing that NRPS enzymes are very sensitive to the precise boundaries of the substituted region, but that activity can be restored to otherwise non-functional subdomain substitutions by modulation of the regional boundaries.</p>