A synthetic biology approach for engineering the biosynthesis of new friulimicin lipopeptide antibiotics

Nonribosomal peptides are secondary metabolites, which include blockbuster therapeutic agents. As a result, there is considerable interest in developing new methods to produce nonribosomal peptide variants with improved properties. Despite this most nonribosomal peptides are too complex for chemical synthesis. However, the Micklefield lab, in collaboration with Biotica, has shown that the reprogrammed, engineered biosynthesis of nonribosomal peptide variants, with altered and improved biological activity is a much more viable approach. Notably, in BBSRC funded projects (36/B12126 & BB/C503662/1) the Micklefield lab elucidated the biosynthetic origins of the lipopeptide antibiotics CDA from Streptomyces coelicolor and developed a wide range of methods for engineering the biosynthesis of lipopeptide variants, which include daptomycin one of the most effective antibiotics currently available [see supervisor sec. & refs therein]. In this new project we will take a major step forward and develop a synthetic biology platform to further develop new biosynthetic engineering approaches. To achieve this we will focus on the biosynthesis and engineering of friulimicin, a lipopeptide antibiotic from Actinoplanes friuliensis. Friulimicin is notable because it is extremely potent and has been shown to inhibit cell wall biosynthesis by a unique mechanism that is distinct from daptomycin, vancomycin and other commercial antibiotics. As such pathogens that are resistant to these antibiotics, which are a major global concern, could potentially be treated with friulimicin. Recently friulimicin entered phase I clinical trials, but was withdrawn due to unfavourable pharmacokinetics (PKs). To overcome this, it will be necessary to generate friulimicin analogues with improved PKs. This is not possible using synthetic chemistry. Moreover, biosynthetic engineering is limited by the fact that the Actinoplanes producer is not well characterised, does not produce significant quantities of the antibiotic and is difficult to manipulate genetically. As a result, development of improved friulimicins has not been possible to date. We will therefore begin by developing a minimal actinomycete host for heterologous expression of the friulimicin biosynthetic genes. We will use S. coelicolor as a host, because unlike the friulimicin producer, it is well characterised, its genome has been sequenced and many genetic tools are available to manipulate this organism. A number of S. coelicolor strains are also available where the host biosynthetic gene clusters have been deleted, providing an ideal minimal background. Accordingly, we will clone the friulimicin gene cluster and establish expression of the biosynthetic genes in S. coelicolor. To optimise production of friulimicin, in S. coelicolor, we will employ the ribosome engineering approach which has been shown dramatically activate the production of secondary metabolites. Having established high levels of friulimicin production, we will then proceed to explore the key unknown steps of friulimicin biosynthesis. This will involve deletion of genes within the biosynthetic gene cluster, followed by structural analysis of the products. Supplementing deletion mutants with putative synthetic intermediates or genes encoding similar enzymes is also informative. The hypotheses generated will then be tested by in vitro studies, which involve overproduction and characterisation of the individual biosynthetic enzymes. Knowledge of the biosynthesis will then be used to explore new methods for engineering friulimicins. For example mutasynthesis approaches will be developed, where key genes are deleted and synthetic intermediate analogues are supplied to the mutants. We will also explore active site modifications of adenylation domains; module or domain swaps of the nonribosomal peptide synthetase; modification of the lipid moiety. New lipopeptides will be characterised (NMR & MS) and properties (MICs & PKs) will be determined.