The design, production and application of targeted antimicrobial peptides to bacterial pathogens of poultry

Widespread use of antibiotics particularly as growth promotors in livestock has been largely responsible for the selection of a high level of antibiotic-resistance within microbial pathogens of livestock, and as such their use is now more restricted. Alternative methods for the control of pathogenic microorganisms are urgently required. An underutilised method of microbial control is the application of anti-microbial peptides (AMPs). AMPs form the basis of innate immune responses to bacteria and other micro-organisms in all classes of life. The most common mode of action of these to bacterial cells is thought to be via a destabilisation of the bacterial cell membrane resulting in cell lysis. AMPs can vary in size from less than 10aas to 60-70aas, their small size enabling rapid diffusion and affording the possibility of chemical synthesis.
The PhD study will apply Next Generation Peptide Phage Display (NGPD) to isolate peptides that bind to the membrane surface of enteric pathogenic bacteria of poultry. Target bacteria will be those that produce significant production losses/costs: Salmonella Typhimurium, Salmonella Enteritidis, Campylobacter jejuni, Clostridium perfringens and avian pathogenic E. coli (APEC). NGPD couples the vast diversity of phage-peptide libraries with the screening power of next generation sequencing coupled with bespoke bioinformatics sorting methods. This technique is much more effective than 'conventional' phage panning in identifying phage-peptides against a specific target. Using NGPD we anticipate being able to identify high numbers of 'targeting peptides' against the bacterial pathogens in parallel. Once targeting peptides have been identified they will be expressed coupled via peptide linkers to known broad specificity AMPs. The fusions will be designed in terms of previous data of AMP efficacy against target bacteria and the linker length will be based on the optimised spacing of AMP-peptides fused to membrane anchors (e.g. Tucker et al., 2018; Cell 172, 618-628). The fusions and 'parent AMP' will then be defined in terms of their MIC against the range of pathogenic bacteria. Their potential toxicity will also be tested against host cell lines. AMPs specific for target pathogens with low 'host-cell toxicity' will be produced as recombinant multi-peptide and individual-peptide constructs and retested for AMP activity in different combinations against mixtures of the pathogen bacteria. The optimal combination of AMPs will be determined by iterative machine learning (Matlab or Python 3) using parameters we identify as important for the performance of AMPs individually (e.g. MIC, stability, efficacy against a range of pathogenic species and strains). In the early stages, we expect there to be multiple optimised cocktails, the efficacy of which will be determined in vitro. The results of these experiments will be used to perform a sensitivity analysis to identify components of the cocktail which contribute the most to performance. This information will then be fed back into the machine learning model in order to optimise the cocktail further. The study will produce cocktails of species-targeted AMPs that have been fully characterised in vitro.