Out of the ground: aerial and exotic habitats of the melioidosis bacterium Burkholderia pseudomallei in grasses in Australia

Abstract

Melioidosis is an emerging infectious disease of humans and animals in the tropics caused by the soil bacterium Burkholderia pseudomallei. Despite high fatality rates, the ecology of B.pseudomallei remains unclear. We used a combination of field and laboratory studies to investigate B.pseudomallei colonization of native and exotic grasses in northern Australia. Multivariable and spatial analyses were performed to determine significant predictors for B.pseudomallei occurrence in plants and soil collected longitudinally from field sites. In plant inoculation experiments, the impact of B.pseudomallei upon these grasses was studied and the bacterial load semi-quantified. Fluorescence in situ hybridization and confocal laser scanning microscopy were performed to localize the bacteria in plants. Burkholderia pseudomallei was found to inhabit not only the rhizosphere and roots but also aerial parts of specific grasses. This raises questions about the potential spread of B.pseudomallei by grazing animals whose droppings were found to be positive for these bacteria. In particular, B.pseudomallei readily colonized exotic grasses introduced to Australia for pasture. The ongoing spread of these introduced grasses creates new habitats suitable for B.pseudomallei survival and may be an important factor in the evolving epidemiology of melioidosis seen both in northern Australia and elsewhere globally.

Figure 1

At field site E which was rich in exotic Tully Grass, soil samples positive for B. pseudomallei are shown together with clusters of Tully Grass (A) or moist soil (B). The circles indicate the results of soil screening for presence of B. pseudomallei over two consecutive dry seasons. A) The cluster of Tully Grass along the southeast edge is shown by a map based on Indicator Kriging which predicts the probability that the threshold value 0 for no presence of Tully Grass is exceeded. B) Probability map for moist to wet soil (approx. >200 mV or >4% vol soil water content).

Figure 2

FISH and confocal laser-scanning microscopy on roots and leaves using a validated set of fluorescently labeled oligonucleotides targeting the 16S rRNA of B. pseudomallei (red); β-Proteobacteria (blue) as positive control and the probe mix non-EUB-338-I, II, III (green) as negative control for nonspecific probe or dye binding. The composite pictures are shown with B. pseudomallei cells in magenta (combination of red and blue). No unspecific green signals were evident for B. pseudomallei signals (see Supplement Figure 3). The orthogonal views (B, E, F, G, H) depict the intra- or extracellular location of B. pseudomallei by showing an internal of three dimensional z-stacks. A) Two B. pseudomallei are seen within a root hair (inside location confirmed by orthogonal view – see Supplement Figure 3) and further β-Proteobacteria among root hairs of Tully Grass whose rhizosphere was inoculated with B. pseudomallei. B) B. pseudomallei in the rhizosphere and inside root cells of Mission Grass from a highly positive field site. C) Various B. pseudomallei and other β-Proteobacteria are shown close to and within stomatal guard cells of a Mission Grass leaf of a highly positive field site. D) Three B. pseudomallei outside of a stomatal opening of a leaf of an introduced pasture grass (Digitaria milanjiana cultivar Jarra) whose rhizosphere was inoculated with B. pseudomallei. E) B. pseudomallei in stomatal guard cells and on the surface of a stomata of a Paspalum leaf collected at field site B. F)–H) B. pseudomallei are seen along vascular bundles of leaves of grasses whose rhizosphere was inoculated with B. pseudomallei. They are at the surface of the lower side of the leaf in a gully-like structure created by the vascular bundles F) Leaf of Sorghum intrans. G)–H) Leaves of Wild Rice.

Figure 3

Semi-quantitative analysis of B. pseudomallei load from day three after inoculation in 90 root or leaf samples with 15 samples per group. Samples were of the time series and plant growth comparison experiments. Load comparison was by comparison of ratios with the denominator being the median growth in all root or leaf samples, respectively. B. pseudomallei specific TTS1 counts were normalized with an internal plasmid control adjusting for differences in DNA extraction and PCR efficiency. R=roots; L=leaves.