Discovery and development of next-generation antibiotics

The rapid emergence and spread of antibiotic-resistant bacteria poses a serious threat to public health worldwide. To tackle these superbugs, the discovery and development of new antibiotics with novel modes of action is critical. Paradoxically, antibiotic drug discovery research endeavours have failed to meet the clinical need for new antibiotics over the past two decades and there are few promising candidates currently in the antibiotic pipeline. The research presented in this thesis was aimed at addressing this urgent global healthcare issue by employing three complementary approaches – semi-synthetic modifications of ‘forgotten’ antibiotics, taxonomy-guided biodiscovery from ‘talented’ novel species of microorganisms and metabolomics-guided discovery – to discover new antibiotics.

Chapter 2: Nidulin is a fungal depsidone antibiotic first isolated from Aspergillus unguis in 1945. Despite nidulin displaying promising antibacterial activity against Mycobacterium tuberculosis and methicillin-resistant Staphylococcus aureus, the compound has received little attention and has not been developed further as an antibiotic lead. Unguinol, a nidulin analogue, was selected as a lead compound to study the nidulin pharmacophore. By applying a semi-synthetic approach, fifteen first-generation unguinol analogues were synthesised. Initial antimicrobial screening revealed that 3-O-benzylunguinol has improved antibacterial activity against S. aureus (MIC 0.2 μg/mL) compared to nidulin (MIC 6.3 μg/mL). Encouraged by this result, a further nineteen benzylunguinol congeners were synthesised to explore the structure activity relationship (SAR) more fully. In vitro antibacterial activities revealed that out of 19 semi-synthetic compounds, two antibiotics, 3-O-(2-fluorobenzyl)unguinol and 3-O-(2,4-difluorobenzyl)unguinol, showed potent antibacterial activity against methicillin-susceptible and methicillin-resistant S. aureus (minimum inhibitory concentration range 0.25–1 μg/mL for both). These two antibiotics are currently being investigated in mouse models for in vivo antibacterial activity.

Chapter 3: An alternative approach to antibiotic drug discovery involves, combining classical taxonomy, chemotaxonomy and total genome mining, to rapidly identify novel species of microbes, which should improve the probability of isolating novel classes of secondary metabolites. By employing this approach, a new species of Australian fungus, Aspergillus albomontis, was identified and prioritised for chemical and biological evaluation. When grown on malt extract agar, A. albomontis produced more than 30 metabolites, including a novel polyketide tetronic acids and two diene-carboxylic acids. A comprehensive biological activity profile was generated using a panel of in vitro antibacterial, antifungal and cell cytotoxicity assays. Given this talented fungus yielded such a diverse class of metabolites, it would be an excellent candidate for genome mining to identify additional silent/cryptic analogues that are not produced under standard cultivation conditions.

Chapter 4: Anthrabenzoxocinones (ABX) are chlorinated polyketides originally isolated from Streptomyces sp. MST-144321. Closer inspection of the metabolic profile of this organism showed that it also produces another two different classes of halogenated metabolites – indolocarbazoles and depsipeptides. Isolation, purification and characterisation yielded six novel metabolites, (+)ABX-G, (–)ABX-K, borregomycins E/F, deschlorosvetamycin A and svetamycin H, and nine previously reported metabolites. Analysis of the genome of the organism revealed that two biosynthetic gene clusters (sve and bab) are necessary for synthesis of these metabolites. Svetamycin is a hybrid metabolite, which shares the piperazic acid, glycolic acid NPRSs and other macrolactam and cyclic peptide BGCs.

Summary: In summary, a combination of semisynthetic modification of existing antibiotic scaffolds, taxonomy-guided and metabolomics-guided discovery and genomics were employed to identify and characterise a suite of novel bioactive lead compounds. Several of the compounds identified in this study showed highly promising antibiotic activity against drug-resistant pathogens and are currently being evaluated using in vivo models of bacterial infection. The molecules isolated also serve as highly diagnostic chemotaxonomic markers, allowing new microbial species to be characterised quickly and efficiently. Clearly, in order to continue to fill the antibiotic pipeline with next-generation chemotherapeutic agents, a “one-size-fits-all” approach is unlikely to be effective. Indeed, the results of this thesis clearly demonstrate that a multifaceted cross-disciplinary approach, drawing on techniques from chemistry, biology, genomics and taxonomy, is a highly effective method of rapidly and consistently discovering potent new antibiotics.