Actinobacillus pleuropneumoniae is a Gram-negative bacterium causing the highly infectious disease porcine pleuropneumonia and is responsible for global financial losses to the swine industry every year. Though the virulence of A. pleuropneumoniae is complex and multifactorial, Apx toxins (ApxI-III) are the major contributing factors that causes lung lesions in pigs. Although vaccines are available to prevent A. pleuropneumoniae infections, they do not give complete protection and typically give protection against the serovars used to prepare the vaccine. Thus, a thorough understanding of gene expression and virulence factors is required to develop broadly protective pleuropneumonia vaccines. This thesis first investigated a novel pathway to prevent and treat pleuropneumonia infection by blocking the interaction between Apx toxins and the host cells. To determine the specific ligands bound by each Apx toxin, glycan array analysis using purified Apx toxins (ApxI-III, both the active and inactive forms e.g. ApxCA and ApxA) was carried out. Expressing both with and without ApxC allowed an assessment of whether this activation is required for interaction with the host glycan receptor. Significant work was needed to optimise overexpression and purification of Apx toxins. Glycan array analysis demonstrated that both ApxI and ApxII toxins bound to very similar glycan structures, such as gangliosides and Lewis antigens. Binding of Apx toxins occurred irrespective of activation by the cognate acyltransferase, ApxC. Interestingly, glycan binding was not observed for the ApxIII toxin, indicating that interaction of this toxin with its already characterised host cell receptor, the CD18 subunit of β2 integrin, likely does not occur via glycan interactions. In recent years, systems known as phasevarions – for phase-variable regulons – have been described in multiple host-adapted bacterial pathogens. Phasevarions result from the rapid and reversible expression of genes encoding cytoplasmic DNA methyltransferases. This results in variable expression of these methyltransferases in a population, with variable genome wide methylation differences within a bacterial population resulting in differential expression of multiple genes via epigenetic mechanisms. In all described cases, phasevarions control expression of current and putative vaccine candidates. The study in Chapters 4 and 5 characterised the phase-variable Type I and Type III R-M systems identified in A. pleuropneumoniae. A study of the distribution of both systems using 210 whole genome sequences demonstrated that the Type I R-M system is present in almost all serovars of A. pleuropneumoniae, whereas different phase-variable Type III mod genes showed colocalisation with specific serovars of A. pleuropneumoniae. This study also demonstrated that individual strains of A. pleuropneumoniae could encode both phase-variable Type I and Type III R-M systems, a phenomenon never before observed. In Chapter 4, phase-variable expression of the Type I R-M system was demonstrated using two prototype strains with a combination of semi-quantitative RT-PCR and a locus specific FAM-labelled PCR assay. The work carried out here also developed a method by which locked strains could be generated that only express a single HsdS variant. Characterisation of the Type III mod genes in Chapter 5 revealed the presence of two distinct phase-variable Type III methyltransferases, modP (which exists in four variants, designted modP1 to modP4) and modQ in A. pleuropneumoniae. The serovar-specific distribution of each of these new mod genes was further confirmed using a second strain collection, comprising 265 strains from the Australian national culture collection. This study demonstrated that these Type III methyltransferases are phase-variable, and that each variant methylates a different DNA target sequence (established using a combination of Pacific Biosciences Single- Molecule, Real-Time (SMRT) sequencing and Oxford Nanopore sequencing). This methylome analysis demonstrated the presence of the first phase-variable cytosine-specific Type III DNA methyltransferase discovered in bacteria. Analysis of changes in gene expression and phenotype influenced by phase variation of each Type III methyltransferase showed that each distinct variant regulates different phasevarions, and has a unique influence on bacterial phenotype, such as antibiotic resistance (modP2), biofilm formation (modP1) and growth rate (modP1). In summary, this thesis has taken the first step to both treating and preventing disease caused by A. pleuropneumoniae by characterising the exact host receptors bound by the major virulence factors ApxI and ApxII and will allow the development of new treatment strategies that can block the interaction between Apx toxins and their cellular receptors. This would allow for more effective treatments and negate the use of antibiotics for the treatment of porcine pleuropneumonia. The characterisation of the phase-variable Type I and Type III DNA methyltransferases described in this study has begun to define the stably expressed antigenic repertoire of A. pleuropneumoniae. This will direct and inform the development of rationally designed vaccines to prevent disease caused by this major veterinary pathogen.