Studies in the ecology and biocontrol of Burkholderia pseudomallei

Burkholderia pseudomallei is a saprophytic bacterium that causes melioidosis, the most common cause of fatal bacterial community acquired pneumonia in the tropics. The disease is endemic in North Queensland and has been studied extensively, yet the ecology of the bacterium remains uncertain. The organism must persist and multiply in the environment to maintain a clinically significant reservoir, This study examined the questions of whether B. pseudomallei was maintained in plants in an endemic region, whether local and imported rice species could be infected in the laboratory, and if so what effect the bacterium had on the growth of the plants and whether biocontrol agents could decrease bacterial loads of B. pseudomallei in a plant infection model. Two rice models of infection with B. pseudomallei were developed, a domestic rice grown in Australia which was then examined by qPCR and fluorescent antibody microscopy and a native (wild) rice found in the local endemic region, also examined using qPCR. Rice cultivar differences in susceptibility to B. pseudomallei and near-neighbor species were identified with another cultivar of rice identified as resistant to any growth inhibition. Two possible agents, a bacteriocin-like compound derived from a Papua New Guinea strain of Burkholderia ubonensis and a bacteriophage cocktail were developed in two different rice species models of infection (one native and one introduced rice species) and a biofilm model. The first biocontrol candidate, a previously reported bacteriocin-like compound extracted from B. ubonensis, was partially purified and chacterised. The second biocontrol agent was a bacteriophage cocktail developed from previously isolated bacteriophages at James Cook University, which were screened for activity against B. pseudomallei isolates of interest, amplified, measured by transmission electron microscopy, quantified and combined. Individual phage samples were all partially effective against the bacterial isolate of interest. However, on combination they completely cleared growth on bacterial lawns. Both the bacteriocin-like compound and the bacteriophage cocktail were equally able to significantly inhibit biofilm formation in a 96 well plate model over 24 hr. The bacteriocin was less effective by 48 hr, while the bacteriophage cocktail, due to self replication activity, maintained its ability to keep the biofilm from growing. As the bacteriocin lost activity over time, it was not used directly in the rice models and B. ubonensis was used instead to identify whether interaction with B. pseudomallei would result in less inhibition of growth. B. ubonensis did not act as an effective biocontrol agent, probably because it was also somewhat inhibitory at the load applied. The bacteriophage cocktail was effective, with significant increases in growth in treatment groups relative to infected groups. The domestic rice model of infection produced partial recovery of infection load while the wild rice model of infection produced complete recovery and a two log drop in bacterial load. The wild rice species likely has a long exposure to B. pseudomallei and has natural defences which, in combination with phage, are able to control the high B. pseudomallei infection. A spatial / botantical analysis of Castle hill in the dry season (a previously described melioidosis-endemic site in Townsville, Queensland) identified particular plant species most commonly found near B. pseudomallei positive soil, however a wet season analysis failed to find B. pseudomallei in roots, although it was found in low levels around the roots and at increased numbers at 10 and 30 cm below ground level. Low exposure and natural resistance of native species to B. pseudomallei, or simply infection below detection limits are probable explanations and it is likely that at this endemic site, plants may play less of a role in survival of B. pseudomallei than does the physiochemical attributes of soil. This study has developed rice models of B. pseudomallei infection and used these to examine the viability of a biocontrol approach with two different types of biocontrol agents. This approach has shown promise in the laboratory. In addition, this study examined the role played by plants in the ecology of melioidosis in an endemic region and how those plants related to B. pseudomallei presence in the soil.