Harnessing the biosynthetic potential of bacteria to produce ribosomally synthesised natural products

Bacteria make an incredible number of chemical compounds that are invaluable for a variety of medical and agricultural purposes, including antibiotics, antifungals, anticancer compounds and insecticides. In fact, the majority of clinically used antibiotics come from soil-dwelling bacteria. This ability to produce these biologically active natural products stems from the evolutionary advantage the molecules provide to the producer. For example, bacteria have evolved the ability to produce powerful antibiotics to kill competing neighbouring microbes. The recent crisis in the rise of multi-drug resistant bacterial infections means that there is a pressing need to discover new antibiotics. We believe that there are many novel antibiotics that remain to be discovered from bacteria, but existing discovery methods are missing many hidden molecules that we call "metabolic dark matter".

These natural products are produced by the action of a series of enzymes (proteins), which are encoded by genes (DNA) in the bacterial genome. Hundreds of thousands of bacterial genomes have now been sequenced. Researchers have developed methods to predict what compounds a bacterium should be able to make based on this genomic data ("genome mining"). This has revealed that many bacteria appear to be capable of producing many more compounds than have been identified. These cryptic compounds may be potent medicines or have other important biological functions. This makes the identification of these pathways and the associated compounds an important research goal.

However, we hypothesise that many important pathways are missed by existing genome mining methods. In this project, we will use a combination of computational, genetic and chemical methods to identify these molecules, understand how they are made and analyse them for biological activity towards clinically and agriculturally important pathogens.

We will focus on discovering new members of a class of natural product called ribosomally synthesised and post-translationally modified peptides (RiPPs). These are made across nature, from bacteria to monkeys, using the same biological machinery that makes large proteins. However, RiPP pathways have evolved to make much smaller natural products that have potent bioactivity. RiPPs include thiostrepton, which is used as an antibiotic to treat bacterial infections in veterinary medicine, nisin, a peptide with broad spectrum antibacterial activity that is used in food processing to suppress bacterial growth, and ziconotide, which is derived from a cone snail RiPP and is used to treat chronic pain in humans.

Ribosomally synthesised and post-translationally modified peptides (RiPPs) are a natural product class that have key ecological roles and significant clinical promise. Multiple RiPPs and their derivatives are used (or are in trials) in medicine, such as thiostrepton, nosiheptide, ziconotide, MOR107 and LFF571. RiPPs originate from a larger ribosomally synthesised precursor peptide that consists of an N-terminal "leader" sequence and a core peptide. The core peptide is post-translationally modified by tailoring enzymes and is then hydrolysed from the leader peptide to yield the mature RiPP.

Despite a requirement to be assembled from proteinogenic amino acids, there is huge structural diversity across the RiPP class. However, there are fundamental challenges associated with the computational identification of RiPP gene clusters, as RiPP biosynthetic pathways lack universally shared features. This contrasts to other natural product classes, whose pathways feature conserved enzymatic motifs that are used in their bioinformatic identification. Therefore, many RiPP gene clusters remain unknown. Recent genomics-led findings of completely new RiPP families with antibacterial, anticancer and antiviral activity highlights how decades of screening efforts have overlooked these important natural products.

To address the challenges associated with RiPP discovery, we developed RiPPER, which represents a new way to identify RiPP gene clusters. In this proposal, we will build on this research to develop tools and resources for the identification of new RiPP gene clusters. This will lead to RiPPER2, a tool that will be made freely available to the natural products community, therefore providing an informatic framework for the discovery of novel RiPPs. We will use microbial genetics, chemistry and biochemistry to characterise new RiPP families (discovered by RiPPER in preliminary work) that we predict will possess antibacterial activity.