Exploring the metabolic diversity of engineered fungal non-ribosomal peptide synthetase-like enzymes for the development of novel antibiotics

A wide range of natural products have important beneficial properties for humans and many of them are produced by fungi. Examples are antibiotics such as penicillin, immunosuppressants such as cyclosporine or the cholesterol-lowering drug lovastatin. These drugs are formed by proteins (enzymes) that are called polyketide synthases (PKS) and non-ribosomal peptide synthetases (NRPS) and the ways in which these enzymes form specific products has been well studied. Penicillin was discovered nearly a century ago and this antibiotic revolutionised the treatment of bacterial infections. Today, its therapeutic use is in jeopardy due to the emergence of multi-drug resistant bacteria and new drugs are urgently needed. This project investigates the exploitation of a new class of fungal enzymes with the potential to produce novel antibiotics. These enzymes are called NRPS-like enzymes and appear to be very widely distributed among mushroom-forming fungi and moulds.

NRPS-like enzymes produce a variety of natural products with potential applications as herbicides and antiviral drugs as well as drugs against cancer and diabetes. Moreover, some of these products display promising antibiotic activity. However, their full potential is as yet unexplored. One reason for this is the limited understanding of the enzyme structure and the lack of knowledge of the mechanism by which these enzymes form their products. This makes it difficult to design enzyme versions that can be used to make products of choice. Therefore, one major aim of this study is to determine the three-dimensional structure of key parts of NRPS-like enzymes to understand how the enzymes make these valuable natural products.

We have found that one of the products with antibiotic potential from an NRPS-like enzyme is effective at inhibiting Mycobacterium tuberculosis, a bacterium that is the main cause of human tuberculosis and that frequently develops extreme multi-drug resistance. As the product from the NRPS-like enzyme attaches to a clinically relevant target in M. tuberculosis, a deeper understanding of the enzyme will help to optimise the generation of even more effective products. Moreover, by combining individual NRPS-like enzymes, a great diversity of novel products can be formed, with widened potential and spectrum for human use. For such enzyme combinations, this project will harness the special features of an NRPS-like enzyme from a lichen. Lichens produce more than 1000 natural products, but their slow growth and the lack of tools for genetic modification hampers their exploitation. We have successfully introduced this lichen-derived NRPS-like enzyme into other fungi that are more suited to this application. While all NRPS-like enzymes characterised so far only produce a single product, this lichen-derived enzyme can generate a variety of products with similarity to the product that inactivates mycobacteria. Therefore, we will characterise the full product spectrum of this lichen-derived enzyme to generate products with increased antibiotic potential. In addition, we will capitalise on its flexibility by making protein combinations (fusions) to generate products that are novel to nature.

Fungi are known for the production of natural products beneficial for humans such as antibiotics, immunosuppressants and statins. Most of these metabolites derive from polyketide synthases and non-ribosomal peptide synthetases with well-characterised reaction mechanisms. We will investigate a relatively new class of enzymes called non-reducing NRPS-like enzymes that are characterised by a three-domain structure consisting of an adenylation, thiolation and thioesterase domain. While the adenylation domain activates a specific substrate, the thioesterase domain condenses two substrate molecules under the formation of various core structures such as benzoquinones, furanones or dioxolanones that interconnect the substrate molecules. One of the metabolites produced by NRPS-like enzymes is atrofuranic acid and preliminary analyses revealed that it efficiently binds to the clinically validated target InhA from Mycobacterium tuberculosis and inhibits its growth. Therefore, atrofuranic acid and derivatives thereof may have a great potential for the development of new antibiotics. For the targeted design of new metabolites from NRPS-like enzymes, an understanding of the chemistry of these thioesterase domains is required and this project aims in the crystallisation and structure elucidation of selected thioesterase domains to model the reaction mechanisms and to identify bottlenecks in the acceptance of different substrates. To use alternative substrates, we will exploit a new NRPS-like enzyme that derives from a lichen mycobiont. This enzyme shows an unprecedented substrate flexibility and is able to condense two different substrates. In addition, the enzyme is not limited to natural, but also activates and uses chemically modified substrates. Combining the flexibility of this adenylation domain with thioesterase domains producing different interconnecting core structures will allow the production of a large scaffold of new metabolites with enhanced antibiotic activity.