Bacterial habitats are usually very diverse and can involve interspecies competition or even competition between similar subspecies since many organisms are trying to utilize the same resources. In the case of kin selection for those of the same subspecies, various methods are employed to self-select for kin including the use of toxin-antitoxin systems, in which all bacteria of the same species in the associated region receive the toxin of interest but those that do not contain the associated antitoxin protein are inhibited by a specific mechanism. These toxic mechanisms include DNase, RNase and pore-forming activity. Additionally, these toxin-antitoxin systems may be deployed by general secretion into the surrounding areas or by means of direct contact, or contact-dependent inhibition (CDI). Characteristic of the various contact-dependent growth inhibition systems found in diverse species is a large toxin protein in which the N-terminal end of the protein is highly conserved and the C-terminal end (CT) is highly divergent. It has been determined that the CTs of these various strains provide the toxin activities of these proteins. Additionally to this there exists immediately downstream of these large toxin genes another gene that encodes an immunity protein. This immunity protein is also highly divergent from other contact-dependent growth inhibitory containing strains, consistent with the idea that they must block activity of the polymorphic CT region of the toxin protein. Thus far, contact-dependent growth inhibition systems have been described in gram-negative bacteria such as E. coli, but none have yet been described in gram-positive bacteria. Bioinformatic analysis has identified a locus in Bacillus subtilis that contains two genes that have similar structural characteristics to that of previously described contact-dependent-growth inhibitions systems.

In this thesis I now describe a new system of contact-dependent growth inhibition that is present in many subspecies of Bacillus subtilis, which is mediated by the toxin protein by the name of wall-associated protein A (WapA). It demonstrates that the toxin activity of these proteins is found at the C-terminal ends of each WapA, which encode tRNases whose activity can be blocked by a protein encoded by the downstream gene of wapA termed wapI. Additionally, while the N-terminal regions of all wapA genes studied so far are highly conserved, the toxin encoding CT regions vary vastly. The highly polymorphic nature of CTs can also be seen in WapI proteins, which can only prevent toxicity of their cognate WapA-CTs. This toxic activity is delivered in a contact-dependent manner, and the toxins themselves target a different set of tRNA species. Bacillus strains that contain differently expressed CTs create a noticeable exclusion interface when allowed to swarm over a semi liquid surface and meet up with the opposing strain, while the interface can be avoided by providing the cognate immunity gene that corresponds to one's opponent. However, the WapA system does not seem to function the same when competing wild type strains, as WapI no longer shows significant protection from other subspecies, indicating that there may be another toxin system in play. This novel toxin-antitoxin system thus offers new insight into means by which gram-positive bacterial species can maintain homogeneity of populations and compete for resources.