Burkholderia pseudomallei strains and other bacterial isolates used in this study are listed in Table 1.

Threonine-basal salt solution with colistin (TBSS-C50) (Galimand and Dodin, 1982) and TBSS-C50-based erythritol medium (Trinh et al., 2019) were prepared as described by Limmathurotsakul and colleagues (Limmathurotsakul et al., 2013) and by Trinh and colleagues (Trinh et al., 2019), respectively. Both media differ in nitrogen and carbon sources. 1.05 mM nitrilotriacetic acid and 50 mM L-threonine in TBSS-C50 was replaced by 18.7 mM NH₄H₂PO₄ and 0.4% (32.75 mM) erythritol as the single carbon source in TBSS-C50-based erythritol medium.

One liter of ACER medium was prepared by dissolving the base components: firstly, the buffer substances 4.75 g (34.9 mM) KH₂PO₄, 2.63 g (15.1 mM) K₂HPO₄ followed by 5 g (0.5%) NaCl were solubilized thoroughly in about 600 mL of destilled water. Following this, 0.12 g (0.49 mM) MgSO₄*7H₂O and 0.02 g (0.14 mM) CaCl₂*2H₂O, each diluted in 50 mL distilled water beforehand, were added to the solution. In the next step, 2.15 g (18.7 mM) NH₄H₂PO₄ and 10 mL of a 1 M NaNO₃ stock (final 10 mM) were added. The base was adjusted to pH 6.3, filled up with distilled water to 876 mL and either sterile filtered or autoclaved. After sterilization, the base was combined with 100 mL sterile filtered erythritol solution of 12% (final 1.2% per liter) and 4 mL of an autoclaved micro salt solution (see below). One liter of the micro salt solution stock contained 11.53 mL H₃PO₄ 85%, 1.49 g ZnSO₄*7H₂O, 0.11 g CuSO₄*5H₂O, 0.63 g MnSO₄*H₂O, 0.15 g Co(NO₃)₂*6H₂O, 0.15 g Na₂MoO₄*2H₂O, and 0.31 g H₃BO₃ in 1000 mL distilled water. To prevent precipitation, 400 μL of 100 mM FeSO₄*7H₂O solved in 2 M HCl and stored at −20°C was added separately per liter of ACER broth (final working concentration 40 μM) directly before use. Also prior to experiments, 20 mL 100x Gibco^™ MEM Vitamin Solution, containing 170 mg NaCl, 2 mg D-Calcium pantothenate, 2 mg choline chloride, 2 mg folic acid, 4 mg i-inositol, 2 mg nicotinamide, 2 mg pyridoxine · HCl, 0.2 mg riboflavin and 2 mg thiamine · HCl, was added. Finally, the colistin sulfate concentration (Carl Roth, Austria) was adjusted to 50 mg/L. Cyclohexmide was also added at a final concentration of 50 mg/L for soil culture experiments. All of the salts and reagents were purchased from either Sigma Aldrich (Austria) or Carl Roth (Austria) suppliers.

Ashdown agar was prepared as described (Ashdown, 1979) containing 10 g trypticase soy broth (Becton Dickinson, United States), 15 g agar, 40 mL 98% glycerol (Fisher Chemical, United Kingdom), 5 mg crystal violet (Sigma-Aldrich, India), 50 mg neutral red (Sigma-Aldrich, United States) supplemented with 5 mg gentamicin (Carl Roth, Germany) per liter.

Regarding growth experiments in liquid culture, bacteria were initially grown on Columbia agar containing 5% sheep blood (BD Biosciences, Austria) at 37°C for 20 h under aerobic conditions. Bacteria were harvested from blood plates and washed twice with phosphate buffered saline (PBS) at 6000 g for 2 min. Cultures were inoculated to an initial optical density of 0.01 at 600 nm (OD₆₀₀). Culture experiments were performed at 40°C either in a volume of 200 μL medium in a Bioscreen C instrument (Labsystems, Helsinki, Finland) under continuous shaking, in 50 mL falcon tubes containing a volume of 10 mL medium or in 125 mL flasks filled with 25 mL of the respective medium. Cultures in falcons and flasks were either shaken at 120 rpm or incubated statically. OD₆₀₀ measurements were performed at the indicated time points. Measurements of pH of bacterial cultures were performed using the HANNA instrument HI98103 (Hanna Instruments Austria).

Twenty soil samples were collected at the end of the wet season in October 2022 in the north-central part of Vietnam from two adjacent sites, a sugar cane field and an uncultivated field separated by a road. These sites were chosen because a previous environmental culture screening in 2017 revealed a high number of B. pseudomallei culture-positive samples in the sugar cane field (Lichtenegger et al., 2021). All soil samples were taken at a depth of 30 cm, with a distance of 5 to 10 m between each other. The steel auger used for collection was cleaned with bottled water and disinfected with 70% alcohol after each sampling point. Approximately 200 g of soil samples collected in plastic zip bags were kept at an ambient temperature and transferred immediately to the Institute of Microbiology and Biotechnology, Vietnam National University, Hanoi. Samples were divided equally and shipped to the Institute of Hygiene, Microbiology and Environmental Medicine, Medical University of Graz, Austria at ambient temperature. Upon arrival in Graz, the soil samples were kept in sealed plastic bags at room temperature and subsequently used for soil enrichment culture experiments (Figure 1). The content of clay, silt and sand was quantified by the pipette method and soil texture was classified by referencing the textural triangle (Staff, 2017) at the Soils and Fertilizers Research Institute (SFRI), Vietnam Academy of Agricultural Sciences. Soil samples were categorized as clay loam (n = 2) and loam (n = 18).

5 g of soil from each sample was homogenized and dissolved in 10 mL PBS for the comparative culture protocol validation. The soil-PBS suspensions were shaken overnight with closed lids at 160 rpm at room temperature. After sedimentation for about 20 min, the upper aqueous phase was collected in fresh tubes and thoroughly mixed. A volume of 500 μL of the aqueous soil suspension was added to 10 mL of ACER, TBSS-C50-based erythritol medium and TBSS-C50 for inoculation. Another 500 μL soil suspension aliquot was kept for PCR analysis and centrifuged for 10 min at 18,213 g, the supernatant discarded and the pellet frozen at −80°C until use. Cultures were incubated with and without shaking at 120 rpm at 40°C for 9 days. Shaken cultures were incubated with closed lids, static cultures with loose caps. Following 48 h of incubation, additional sub-enrichments of all media were started: A volume of 500 μL of the 48 h enrichments was transferred to 10 mL of the respective fresh medium and further incubated statically for 7 days (Figure 1). A volume of 500 μL of the enrichment cultures was frozen on the last day of each enrichment culture either as a pellet, as described above for qPCR analysis, or as glycerol stock (25% glycerol) at −80°C for subsequent subcultures on agar.

The DNA extraction from bacterial pellets was performed by using the NucleoSpin^® Microbial DNA Kit from Macherey-Nagel with slight modifications as described previously (Wagner et al., 2023). Quantitative PCR was performed using a probe-based PCR real-time assay specific for a TTSS1 gene region of B. pseudomallei (Novak et al., 2006; Trinh et al., 2019).

Glycerol stocks of soil enrichment cultures were thawed, mixed thoroughly and 100 μL of tenfold serial dilutions plated in duplicate on Ashdown agar. Plates were incubated at 40°C for 4 days followed by 1 day at room temperature. Morphologically suspicious B. pseudomallei colonies were identified and confirmed by PCR, targeting the B. pseudomallei specific sequence of the TTSS1 gene (Novak et al., 2006).

Statistical analyses and calculations were performed using GraphPad Prism software version 9.2.0 (GraphPad Software, San Diego, CA, United States). The comparison between optical densities at respective time points and media compositions was done with the Wilcoxon matched-pairs signed rank test and Friedman test followed by Dunn’s multiple comparisons test. The growth of eight B. pseudomallei strains in ACER, TBSS-C50, and TBSS-C50-based erythritol media was analyzed with Fisher’s exact test. C_T values and colony forming units (CFUs) of different enrichment cultures of the same soil sample suspension were analyzed as paired data with the nonparametric Friedman test followed by Dunn’s multiple comparisons test. Missing C_T values were assigned to an arbitrary C_T of 40 to account for the need of complete data sets for Friedman analysis. The interrelation between cultivated B. pseudomallei bacteria on Ashdown agar and respective C_T values was calculated with the Spearman’s rank correlation. Median cycle threshold (C_T) values of TTSS1-qPCR data from soil enrichments are indicated in the text in parentheses with the lower quartile q1 and the upper quartile q3.

We recently demonstrated the highly selective growth of B. pseudomallei strains in TBSS-C50-based erythritol medium (Trinh et al., 2019). We aimed here to optimize medium parameters to further improve growth and, at the same time, preserve selectivity. Since there is evidence that the occurrence of the genus Burkholderia (Stopnisek et al., 2014) and particularly B. pseudomallei (Chen et al., 2003; Kaestli et al., 2007; Palasatien et al., 2008; Kaestli et al., 2009) is associated with acidic soil conditions, we, firstly, investigated the influence of the medium pH on the growth of various B. pseudomallei strains (Table 1) in a pH range from 5.5 to 7.2 (Figure 2A; Supplementary Figure S1).

All strains benefited in early growth during the first 24 h from an initial acidic medium pH down to 6.0 (Figure 2A; Supplementary Figures S1, S2). After 72 h of enrichment, we still saw a significant increase in growth for pH 6.3 compared to 6.0 and 6.5 (Supplementary Figure S2). However, we observed that in cultures of lower initial pH, this positive trend was reversed with ongoing incubation time (Figure 2A; Supplementary Figure S1). We hypothesized that this later growth stagnation of initially acidic cultures was caused by the continuing acidification of the culture. We increased the buffer capacity from 13 to 50 mM to maintain the advantage of a shortened lag phase but stabilize the growth in acidic culture medium over time. As depicted in Figures 2B–D, higher buffered cultures with an initial pH of 6.3 (Figure 2C) reached higher optical densities than cultures of pH 6.0 (Figure 2B) and 6.5 (Figure 2D) within 96 h and comparable optical densities at 144 h compared to pH 7.2 (Figure 2A). Therefore, a 50 mM phosphate buffer concentration and a pH of 6.3 were chosen for subsequent medium optimization experiments.

We next tested different erythritol concentrations in 50 mM phosphate-buffered TBSS-C50-based erythritol medium at pH 6.3. Increasing the erythritol concentration up to 1.2% led to more than twice higher optical density compared to erythritol concentrations of 0.4% (Figure 3A), meanwhile a further increase in the erythritol concentration above 1.2% led to a reduction in growth. The lag phase was further reduced by the supplementation of vitamins (Gibco^™ MEM Vitamin Solution). The optical densities had already more than doubled by day two of the incubation in media enriched with vitamins in 1:50 dilution, and on day 3 more than four-fold compared to growth in 1.2% erythritol medium without vitamins (Figure 3B). We therefore continued our medium optimization with 50 mM buffered medium at pH 6.3 containing 1.2% of erythritol and MEM vitamin solution diluted 1:50.

B. pseudomallei encounters hypoxic or even anaerobic conditions in its natural soil habitat (Tiedje et al., 1984; Liesack et al., 2000; Yamauchi et al., 2000; Angel et al., 2012). Therefore, at least a fraction of the soil-dwelling B. pseudomallei population might be metabolically adapted to a limited oxygen supply at the time when a soil sample is subjected to an enrichment culture. Furthermore, if static cultures are applied instead of shaken cultures, the availability of oxygen for cellular respiration is reduced. Since B. pseudomallei is capable of using nitrate as a terminal electron acceptor for respiration (Yabuuchi et al., 1992), we investigated the growth of B. pseudomallei strain K96243 in erythritol medium with the addition of nitrate. Prior to these experiments, the impact of different nitrate concentrations on B. pseudomallei shaken cultures was tested and 10 mM was chosen for subsequent experiments (Supplementary Figure S3). Figure 4 shows experiments with and without the addition of nitrate under shaken (Figure 4A) and static conditions (Figure 4B). Growth was generally clearly higher in shaken (Figure 4A) compared to static cultures (Figure 4B). Higher cell densities were observed in static enrichments in the presence of 10 mM sodium nitrate (Figure 4B). Since the addition of nitrate improved the growth under static culture conditions without any significant inhibitory effect, we included 10 mM nitrate in the medium formula from this point onwards.

Finally, we reduced the sodium chloride concentration in the medium base from 1 to 0.5% because some strains grew better at sodium chloride levels below 1% while none showed impaired growth at the reduced salt amount (Supplementary Figure S4). In summary, our newly composed acidic erythritol medium, designated ACER medium differs from our previous TBSS-C50-based erythritol medium in a lower pH of 6.3, an increased concentration of 50 mM phosphate buffer, an erythritol concentration raised to 1.2%, sodium chloride reduced to 0.5%, added vitamins in a dilution of 1:50 and the addition of 10 mM nitrate.

We, finally, validated the new ACER medium with eight B. pseudomallei strains (Figure 5; Supplementary Figures S5, S6) and compared their growth to TBSS-C50-based erythritol medium and TBSS-C50 under shaking (Figure 5A; Supplementary Figure S5) and static (Figure 5B; Supplementary Figure S6) conditions. After 144 h of shaking and static incubation, seven out of eight strains reached higher optical densities in ACER medium compared to TBSS-C50-based erythritol medium (p = 0.02) and all strains grew better compared to TBSS-C50 (p = 0.003). The selectivity of ACER medium was confirmed in experiments with other soil-dwelling species, namely, Burkholderia thailandensis, B. cenocepacia, B. stabilis, B. multivorans, B. cepacia, and Cupriavidus gilardii, which did not show any growth in ACER medium (Supplementary Figure S7).

To finally test the newly developed ACER medium for its potential to increase the detection and recovery of B. pseudomallei from soil samples, we validated ACER medium in comparison with TBSS-C50-based erythritol medium and TBSS-C50 with twenty samples collected at the end of the wet season in October 2022 in the north-central part of Vietnam. Comparing 9 days continuous static culture of ACER medium (C_T = 14.12; q1 = 13.62, q3 = 15.53) with TBSS-C50-based erythritol medium (C_T = 16.30; q1 = 16.07, q3 = 17.41, p = 0.0002) and TBSS-C50 broth (C_T = 17.19; q1 = 16.79, q3 = 18.66, p < 0.0001) by qPCR clearly revealed a superior enrichment of B. pseudomallei in ACER medium (Figure 6). Interestingly, the comparison of static and shaken 9 days ACER cultures (C_T ^{9d, shaken} = 15.87; q1 = 15.27, q3 = 17.80) revealed no improved but a slightly lower enrichment of B. pseudomallei in the latter and a less clear advantage of ACER medium (Figure 6). We furthermore investigated the effect of a static ACER two-step culture with a transfer step after 2 days into new medium followed by 7 days of incubation (Figure 6). Again, we detected the highest enrichment with ACER medium (C_T ^{2 + 7d, static} = 15.06; q1 = 14.84, q3 = 18.82) compared to TBSS-C50 (C_T ^{2 + 7d, static} = 17.46; q1 = 17.02, q3 = 19.93, p < 0. 0001) and TBSS-C50 based erythritol medium (C_T ^{2 + 7d, static} = 18.06; q1 = 17.01, q3 = 20.25, p = 0.003). However, in this setting, the two-step procedure was not superior to the continuous static enrichment for 9 days (Figure 6).

We next cultivated qPCR positive soil enrichments from all three media after 9 days of static cultures on Ashdown agar (Figure 6, respective sample sets highlighted with a grey box). Sixteen soil samples with CT values below 30 in all three media were culture positive (Figures 7A,C). The remaining culture-negative samples were either PCR-negative or had CT values above 30 (Figure 6) of which the latter did not lead to the recovery o B. pseudomallei from agar. This is in accordance with our previous study, where no B. pseudomallei growth was obtained on Ashdown agar from any TBSS-C50 or TBSS-C50-based erythritol medium culture with C_T values above 30 (Trinh et al., 2019) 30.7 times higher CFU counts were recovered from ACER enrichments (median CFU = 2.27 * 10⁷, q1 = 8.86 * 10⁶, q3 = 4. 70 * 10⁷) compared to TBSS-C50 (median CFU = 7.4 * 10⁵, q1 = 2.74 * 10⁵, q3 = 7.61 * 10⁶; p < 0.00001) and 5.4 times higher TBSS-C50-based erythritol medium cultures (median CFU = 4.2 * 10⁶, q1 = 2.68 * 10⁶, q3 = 1.98 * 10⁷; p = 0.005) (Figure 7A). The comparison of B. pseudomallei CFUs with the corresponding C_T values from all enrichments (Figure 7B) revealed a strong negative correlation between the C_T values and the recovery of B. pseudomallei CFUs on Ashdown agar (r_S = −0.68, p < 0.0001). Figure 7C depicts the recovery of viable cells from the three different culture media for single soil samples together with the corresponding C_T values. In all 16 samples, ACER led to the highest CFU numbers of B. pseudomallei on agar plates. The increase in the recovery of viable cells from single samples cultivated in ACER medium compared to TBSS-C50-based erythritol medium ranged from 1.19 fold to 18.99 fold. The increase in ACER medium compared to TBSS-C50 ranged from 3.17 fold to 228 fold.