Investigating natural and novel PKS-NRPS genes

Many organisms produce secondary metabolites, which are compounds, often with complicated structures, that are not necessary for the normal growth and development of the organism that produces them. Polyketides (PKs) are a class of secondary metabolites that are produced by many bacteria and fungi. Some PKs are toxic and must be avoided, but others have useful pharmacological properties and are used in human and veterinary medicine and agriculture. These include various antibiotics used in combating infection, and the statins, which reduce cholesterol synthesis. This project seeks to shed light on the way in which PKs are made by filamentous fungi. Many genes, encoding many enzymes, contribute to specifying a particular PK structure. Foremost among these enzymes is the polyketide synthase (PKS), which assembles a basic carbon backbone, and this backbone is modified by the activities of other enzymes. This project concerns a particular class of PKS proteins, ones that have an integral non-ribosomal peptide synthase (NRPS) module, whose function it is to add a specific amino acid to the PK backbone. An ultimate objective of the work is to be able to manipulate the genes involved in PK synthesis in order to generate novel chemicals with desirable biological properties. To do this requires understanding of the various enzymes and functional domains within complex enzymes like PKSs and PKS-NRPSs. Such understanding can be obtained by looking at the activities of parts of the system in isolation and by mixing and matching components to see what results. Because PKS-NRPSs are very large multifunctional proteins it is very difficult to manipulate the genes that encode them. In preparation for this project we have developed a system that simplifies both the construction of large genes from gene fragments and the transfer of the final product into the plasmid vectors in which they are required for further analysis. The system also simplifies the construction (and transfer) of chimaeric genes containing DNA from different sources. This will enable us to investigate various aspects of the activities of PKS-NRPSs. First, we will be able to discover the function of NRPS modules from several genes / some of which are only known to exist from genome sequences. We will attach NRPS modules to a PKS that has had its own NRPS module removed and see what amino acid gets added to the PK backbone. Using the same mix-and-match approach, but on a finer scale, we will then dissect the NRPS modules to discover where the amino-acid specificity resides. We would hope to be able to progress from this experimental approach towards making predictions of amino-acid specificity from scrutiny of DNA sequences alone. Activity of the chimaeric genes in a host organism that does not have the tailoring enzymes will produce novel products that are not usually encountered in nature. We will produce additional novel compounds by putting the chimaeric genes directly into the gene cluster from which the original PKS-NRPS was isolated. All novel compounds have the potential for use in biological interventions directly or after rational modification. In another series of experiments we will compare pairs of very similar PKS-NRPSs to discover how differences in structure affect their function. One pair differs only in the presence or absence of a short protein segment that may confer an extra activity on the enzyme that has it. We will look for differences in structure of the products of expressing the gene that has the additional segment and the same gene with segment removed. Another pair is known to produce products that differ only in the length of the PK chain. By comparing what we predict to be very similar genes we expect to identify sequences responsible for 'programming' the number of cycles that the enzyme goes through (which determines chain length), and we will test this by converting the gene from a pentaketide producer to a hexaketide producer.

Fungal polyketides (PKs) are important secondary metabolites, many of which have useful pharmacological properties that are exploited in human and veterinary medicine and agriculture. Their synthesis is encoded by gene clusters encoding a polyketide synthase (PKS), a large, multifunctional, iterative enzyme responsible for assembling the carbon backbone, and various tailoring enzymes that modify the initial PK product. Some PKS have integral non-ribosomal peptide synthase (NRPS) modules that add an amino acid to the PK backbone prior to tailoring. Understanding the mechanisms underpinning PK synthesis could lead to the rational design of new drugs by combinatorial genetics, but the huge size of PKS and PKS-NRPS genes has hampered progress. By combining homologous recombination in yeast and the Gateway system for gene transfer we have developed a system for the construction of very large native and chimaeric coding regions and their simple transfer to expression vectors. We will use this system to investigate mechanistic features of fungal PKS-NRPS proteins. To discover the amino-acid specificity of several NRPS modules we will delete the NRPS module from the fusA gene responsible for fusarin C biosynthesis and replace it with heterologous NRPS modules. Expression in heterologous hosts will produce novel, 'untailored' compounds, whereas recombination in situ in Fusarium venenatum will place the hybrid gene within the fusarin C gene cluster. To refine this work we will make domain swaps within the fusA NRPS to determine where specificity resides. To investigate chain length programming we will compare the PKS-NRPSs responsible for pentaketide tenellin and hexaketide bassianin production. By exchanging sequences within the tenA gene we aim to increase activity by one cycle resulting in bassianin production. To investigate protein flexibility we will compare the activities of the fusA gene with and without a sequence lacking from an otherwise identical PKS-NRPS.