Bacteriocins are ribosomally synthesised antimicrobial peptides, produced by many bacterial genera that display potent activity against closely (narrow spectrum) or distantly related (broad-spectrum) bacteria. Bacteriocins, produced by Lactic Acid Bacteria (LAB) that are natural constituents of fermented foods, are ideal natural preservatives to control food spoilage/pathogenic bacteria in minimally processed foods. In addition to their role as food preservatives, bacteriocins have potent activity against medically significant pathogens and are considered attractive alternatives or adjuncts to antibiotics, due to their inherent heat stability, potency at nanomolar scale, resistance to proteases and low levels of acquired resistance in commercial applications. Overall, bacteriocins are versatile antimicrobials with huge potential for use as biopreservatives, antibiotic alternatives, health promoting gut modulators and animal growth promotors. The aim of this thesis was to identify, purify and characterise novel bacteriocins from microorganisms isolated from a wide range of niches, with a view to expanding the number of bacteriocins currently available and exploring novel structures and activities. In this respect, Chapter 2.1 describes the discovery a novel nisin A variant, nisin H, produced by a porcine gut isolate Streptococcus hyointestinalis DPC6484. Nisin H differs from nisin A at five amino acid positions and is an intermediate between naturally occurring nisins of lactococcal and streptococcal origin. The operon encoding nisin H is noteworthy by virtue of the absence of an equivalent of nisI that encodes an immunity protein that protects the cell from its own bacteriocin. This is the first report of natural nisin variant production by an intestinal isolate of streptococcal origin and may confer an advantage to the strain by allowing it to dominate its environment, fight infection or signal the immune system of the host. In a subsequent chapter another natural variant is characterised in the form of nisin J, produced by a human skin isolate Staphylococcus capitis APC2923. Nisin J is more dissimilar to nisin A than nisin H with nine amino acid changes, six of which are unique, and an extra amino acid making it the first nisin variant to contain 35 amino acids. Interestingly, the operon lacks both nisI (immunity) and nisRK (regulatory) equivalents. Nisin J, like nisin A and H, displays activity against a wide number of genera and represents the first natural nisin variant from staphylococci and the first nisin producer from human skin, suggesting a role in competitive colonization for producing organisms. The natural nisin variants described above (nisin H and J), in addition to nisin P produced by Streptococcus agalactiae DPC7040, are all produced by non GRAS strains and are therefore limited in their potential industrial applications. The recent increase in the prevalence of antibiotic resistant pathogens makes it important that all bacteriocins regardless of the producing organism are explored as antibiotic alternatives. As these lantibiotics are gene encoded, bioengineering (Chapter 3.1) was used to enable recombinant expression of peptides naturally expressed by non-GRAS organisms in a host derived from safe origins. Specifically, the Nisin A promotor and nisin A leader sequence were fused to nisin H, J or P structural genes and successfully expressed in the GRAS strain L. lactis NZ9700, demonstrating that the L. lactis production, transport and modification machinery can produce fully functional nisin variants from significantly different genetic backgrounds. In Chapter 4, Bactofencin A produced by Streptococcus salivarius DPC6502 was discovered following a porcine gut mining study. It is a 22 amino acid, class IId bacteriocin that displays activity against Staphylococcus aureus and Listeria monocytogenes. Structurally, it consists of a positively charged N terminus that we propose could bind to the negatively charged cell surface. The small bacteriocin cluster also encodes a DltB homologue that may well be responsible for immunity through D-alanylation of teichoic acids. In order to probe structure/function relationships in bactofencin A, a library of synthetic bactofencin A peptide variants were synthesized. Substituting cysteine residues significantly reduced activity confirming the importance of the disulphide while sequential removal of the positively charged N terminal resulted in a decreasingly active peptide. Substituting each amino acid for alanine revealed that residues 9-17 within the loop were more affected by substitution, suggesting this region contributes significantly to the potency of the bacteriocin. In Chapter 5, bactofencin A was shown to enhance nisin bactericidal activity and reduce the overall frequency of resistance. Interestingly, these studies highlighted the relatively slow or delayed mode of action of bactofencin A. The last two chapters (Chapters 6.1 and 6.2) again focus on the discovery of two novel bacteriocins, namely formicin and actifensin. The first of these, formicin, is a novel bacteriocin that extends the class of two peptide lantibiotics. It was purified from Bacillus paralicheniformis APC1576, a mackerel intestine isolate. Compared with other two component lantibiotics, formicin is most similar to haloduracin and consists of a very hydrophilic Alpha peptide with a charge of +2 whereas the Beta peptide is negatively charged. Formicin displays activity against a broad range of Gram-positive bacteria including clinically relevant pathogens. The second bacteriocin is actifensin a 4091 Da, broad spectrum, Class IId bacteriocin containing three disulphide bridges with more than 50% similarity to eukaryotic defensins that we propose represents a new subclass of bacteriocins. It is produced by Actinomyces ruminicola, isolated from sheep feces. A pangenomic screen of available Actinomyces spp. revealed the presence of very diverse actifensin homologues in 29% of genomes examined, suggesting that production of actifensin like bacteriocins is a common trait. This new class of bacteriocins may provide a template to design new broad-spectrum antimicrobials for treatment of human and animal infections. The developments described in this thesis can be used to contribute to increased commercialisation of bacteriocins in both food systems and human and animal medical treatments.