Figure 1

Figure 1. Sample collection for detection of Mycobacterium ulcerans DNA in Buruli ulcer–endemic area, northern Queensland, Australia. Environmental samples were collected in the Daintree River basin during September (yellow) and October (green) 2013....

We collected 102 environmental samples (55 from soil/mud/vegetation, 35 from insects or small insect pools, and 12 from animal excreta) in September 2013 from different locations within the Daintree River basin, in which BU cases were reported during the outbreak in 2011. Global positioning system coordinates were recorded for all sampling locations (Figure 1). Specimens were collected in sterile plastic containers and shipped to the Victorian Infectious Disease Reference Laboratory (Melbourne, Victoria, Australia) for PCR analysis.

We extracted DNA by using the FastPrep Instrument (MP Biomedicals, Solon, OH, USA) as described (4). We used the FastDNA Kit (MP Biomedicals) for insect samples and the FastDNA Spin Kit (MP Biomedicals) for soil, mud, vegetation, and feces samples. We analyzed DNA extracts by using semiquantitative real-time PCRs optimized for detection of M. ulcerans in environmental samples (9). We first screened all extracted DNA samples for M. ulcerans insertion sequence element IS2404. Subsequently, we analyzed IS2404-positive samples in a second real-time PCR to detect 2 additional regions in the genome of M. ulcerans: IS2606 and a sequence encoding the ketoreductase B domain of the mycolactone polyketide synthase genes.

Figure 2

Figure 2. Sample locations for detection of Mycobacterium ulcerans DNA in Buruli ulcer–endemic area, northern Queensland, Australia. A) Pond where IS2404-positive soil sample was collected at first sampling time in September 2013; B)...

Of the 102 samples, 5 (1 soil specimen, 2 bandicoot [Isoodon macrourus] feces samples, 1 sample of an individual mosquito, and 1 pool of 2 mosquitoes) were positive for IS2404. Although 3 of the 5 specimens did not contain sufficient amounts of DNA to identify IS2606 and the ketoreductase B domain of the mycolactone polyketide synthase genes, as indicated by the high cycle threshold (C_t) values for multicopy IS2404 (Table), these markers were detectable in the other 2 samples (bandicoot feces and the pool of 2 mosquitoes) (Figure 1). The IS2404–positive soil sample and the 2 bandicoot feces specimens were collected at the same location (Figure 1), close to a small pond (Figure 2, panel A), and 2 positive mosquito samples were collected at Wonga Beach (Figure 1).

As reported (9), analysis of difference in real-time PCR cycle thresholds between IS2606 and IS2404 (ΔC_t [IS2606 – IS2404]) enables differentiation between strains of M. ulcerans known to cause BU in humans and other mycolactone-producing mycobacteria strains that contain IS2404 but have fewer copy numbers of IS2606 and are not known to cause disease in endotherms. Thus, DNA in bandicoot feces could be attributed to the M. ulcerans genotype known to cause BU (Table). Conversely, the mosquito pool contained DNA of a closely related M. ulcerans subspecies that had a low copy number for IS2606 (Table), which is usually not associated with disease in endotherms.

The location at which the bandicoot specimen containing M. ulcerans DNA was collected was situated in a nature refuge. In October 2013, we performed a followup environmental study that focused on environmental samples, predominantly animal excreta, collected in this refuge. Of 18 soil/mud/vegetation specimens, 12 insects/insect pools, and 74 animal excreta samples collected, only 1 sample, a bandicoot feces specimen found at almost the same location as the M. ulcerans–positive bandicoot feces specimens collected during the first sampling (Figure 1), was also positive for all M. ulcerans DNA markers tested (Table).

Motivated by increased evidence for the role of M. ulcerans–infected possums in the ecology of human BU in southeastern Australia (4,5), we searched for a similar animal reservoir of M. ulcerans in a BU-endemic area of northern Queensland. Detection of M. ulcerans DNA in bandicoot feces at 2 sampling time points separated by 4 weeks suggests that small mammals might be a potential reservoir for this pathogen. Our data provide a basis for future investigations, which should include a survey of the local animal population for lesions of BU. Similar studies have not identified a mammalian animal reservoir in BU-endemic regions of West Africa (10). Therefore, investigations should be conducted to determine whether at least some marsupial species are more susceptible to M. ulcerans infection than other mammals.

In Far North Queensland, the Daintree River obtains its waters from the mountainous rainforest region northwest of the small town of Mossman and flows into the sea at Cape Tribulation. The wet season is November/December–April, and the dry season is May–October/November. We encountered different environmental conditions at the 2 sampling time points. In September, water bodies, including creeks, small lakes, and swamps, were filled with water (Figure 2, panel A). However, water levels were comparatively low at the end of the dry season in October; some water bodies had even dried up (Figure 2, panel B). Samples from other water bodies, such as creeks (Figure 2, panel C) or biotopes near houses (Figure 2, panel D) in which the presence of M. ulcerans was suspected, all showed negative results for IS2404.

Because outbreaks of BU in Far North Queensland were historically related to heavy rainfalls and flooding, future studies in this region should be performed during or shortly after the rainy season. Although mosquitoes seem to be involved in the ecology of BU in Victoria (5,11,12), the mosquito sample positive for M. ulcerans in this study did not have a bacterial genotype that is known to commonly cause disease in endotherms.