2. Materials and Methods
During the course of an extensive ecological–epidemiological study, samples of peat, peat substrates, soil, animal faeces, wastewater, and watercourse sediments were collected for assessment as potential sources of pathogenic and NTM. Large quantities of samples were collected per day, and consequently an optimised method to process samples and test for the presence of mycobacteria was required to accommodate such a large intake at once. Although extensively used for clinical samples, decontamination with a sodium hydroxide solution requires a long incubation step that was not suitable for these types of samples, where the processing for this study was required to be completed within 24 h of collection. In the relevant literature there was no consensus as to the most efficient way to process environmental samples for mycobacterial culture in the required time constraints [
25,
26].
2.1. Samples’ Collection
A total of 787 samples of different environmental matrices were collected in the territory of Moravia as a part of the Czech Republic in period from January 2019 to March 2021. These samples were collected as part of a study to monitor the spread of mycobacteria and evaluate the species spectrum of mycobacteria in the natural environment, especially in karst and protected landscape locations, as well as area subject to mining activities. These samples were divided in to three main clusters:
Cluster S (274 soil and water sediment samples of rivers and water bodies and sewage sludge from wastewater treatment plants collected in man-made systems, settlements and municipalities, industrial and mining areas, artificial bathing tanks, and related and adjacent rivers and streams).
Cluster R (251 samples of raw and processed peat originating from forest peatlands, gardens, and commercial horticultural products).
Cluster E (262 faeces and products from animals including birds, bats, and other mammals).
2.2. Pre-Treatment Procedure
Material weighing from 5 to 10 g (with a maximum volume of 15 mL) was resuspended in 20 mL of distilled water containing 1% Tween 80 and 20 glass beads (2 mm diameter). Suspension was homogenised thoroughly by vortexing for 1 min, followed by shaking at 300 oscillations/min for 15 min. The homogenised samples were centrifuged at 500 revolutions per minute (rpm) for 10 min. The supernatant was collected and centrifuged at 4300 rpm; this supernatant was then discarded, and the pellet was subsequently used.
2.3. Decontamination with NaOH and TDAB
Fresh decontamination solution was prepared containing 4% NaOH (Lach-Ner, Neratovice, Czech Republic) and 0.5% TDAB (tetradecyl-trimethylammonium bromide; Duchefa Biochemie B.V, Haarlem, The Netherlands), stable for 14 days at 4 °C. The pre-treated pellet was resuspended in 10 mL of decontamination solution, vortexed, and then homogenised on a shaker for 15 min. After centrifugation for 20 min at 4300 rpm, the supernatant was discarded, and the pellet resuspended in 15 mL of distilled water; importantly, this decontamination step did not exceed 40 min. The homogenisation and centrifugation steps were repeated, and the pellet was resuspended in 2.5 mL of phosphate buffer (pH 6.8) and thoroughly mixed and immediately submitted for culturing. Four Löwenstein–Jensen media slants were inoculated with 0.2 mL (0.8 mL total) of processed sample and placed in pairs in a thermostat at 30 and 37 °C and left to incubate for 12 weeks, with growth assessments performed weekly. A 0.8 mL aliquot of processed sample was taken in a separate 1.5 mL Eppendorf tube for qPCR.
2.4. Isolates Identification
Suspected mycobacterial isolates were examined macroscopically and subsequently by microscopy with Ziehl–Neelsen staining. Preliminary identification was made by GenoType
® Mycobacterium CM/AS (Hain Lifescience GmbH, Nehren, Germany). Slightly contaminated and suspected mixed isolates were re-cultured, and secondary isolates were subjected to further identification. The DNA of isolates was extracted and used as template for PCR amplification of the 16srRNA and
hsp65 genes using the universal bacterial primers 5′CCT ACG GGN GGC WGC AG3′ and 5′GAC TAC HVG GGT ATC TAA TCC3′ of the V3 and V4 variable regions [
54] and
Mycobacterium hsp65 primers 5’ACC AAC GAT GGT GTG TCC AT3’ and 5’CTT GTC GAA CCG CAT ACC CT3’ [
55], respectively. The resulting amplicons obtained by both of the methods were purified and sequenced by Eurofins (Ebersberg, Germany). Identification of mycobacterial species was performed by BLAST analysis. Isolates belonging to
M. avium complex members were further identified by the PCR method for the detection of the IS
901 amplicon specific for
M. avium ssp.
avium and the IS1245 amplicon specific for
M. avium ssp.
hominissuis [
37].
Evaluation of accompanying microflora (contamination): All cultures were evaluated microscopically, and massive growth after one day of cultivation was inoculated with a loop by cross-smearing on blood agar and further cultured. From the solid media, which were degraded by the proteolytic activity of the bacteria to a fluid, 100 μL was carefully removed with a Pasteur pipette and inoculated with a line and smeared on blood and Endo agars (for the selection of G- microbes). The media were cultured at 30 and 37 °C. Growth was assessed after 24 h and the culture was closed after 5 days. Individual colonies were identified by MALDI-TOF (MALDI Biotyper, Bruker-Daltonics Billerica, MA, USA).
2.5. qPCR Method
This qPCR method was selected as a reference method for the evaluation of mycobacterial presence using the commercial Real-time PCR Z-Path-Mycobacterium_spp detection kit for Mycobacterium (Primerdesign Ltd., Camberley, UK). DNA was isolated from 0.8 mL of decontaminated sample by centrifugation at 14,000 rpm/min, discarding the supernatant and then processing the pellet with the DNA E.Z.N.A.® Soil DNA Kit (Omega Bio-tek, Norcross, GA, USA) according to manufacturer’s instructions. A total of 0.5 µL of DNA isolation product was added to a Precision PLUS 2× qPCR Master Mix (Primerdesign Ltd., Camberley, UK), and the qPCR reaction was run with a CFX96 real-time PCR detection system (Bio-Rad Laboratories, Hercules, CA, USA) using the following thermocycler conditions: enzyme activation at 95 °C for 2 min, 50 cycles at 95 °C for 10 s for denaturation and at 60 °C for 60 s for aneling and data reading.
2.6. Viability Testing by PMA Treated DNA
An alternative approach using a PMA qPCR method was tested to determine the viability of mycobacteria after the decontamination process. A total of 100 μL of raw and decontaminated suspensions (after neutralisation) were treated with suspension of 50 μM PMA (Biotum Inc. Hayward, CA, USA) diluted in 20% dimethyl sulfoxide (DMSO; Sigma-Aldrich, Burlington, MA, USA) and incubated in the dark for 5 min; then, the mixture was exposed to light from a 650 W halogen bulb for 2 min [
48]. Cells were lysed, DNA was isolated, and qPCR was performed as described previously (
Section 2.5) to determine the viability of cells. Viability determination is based on two separate qPCR reactions.
2.7. Statistical Analysis
Calculation of the average number of colony-forming units (CFU) per 5 g of solid sample: The average CFU per media was first calculated by dividing the CFU count by the number of medium with visible growth. If the strain did not grow at a certain temperature, the results were calculated only from the data of those incubation temperatures at which the strain was able to grow. CFU per gram of solid sample was then calculated using the appropriate dilution factors. For statistical processing, soil colony counts were generalised to the following ranges: 1–10 CFU, 11–100 CFU, and >100 CFU (too numerous to count = TNTC) for continuous colonies and confluent growth (
Scheme 1).
Curves for quantitative PCR were derived from the appropriately diluted internal standard included in a qPCR kit. Results are represented as relative fluorescent units (RFU), approximated according to a measured standard by CFX-manager 2.0. software (Bio-Rad Laboratories, Hercules, CA, USA). Data analysis was performed using statistical software Statistica 13.2 (StatSoft Inc., Tulsa, OK, USA). p-values less than 0.05 were considered statistically significant. For multiple comparison tests, Bonferroni adjustment of p-values were used. Simple summary calculations were performed in the MS Excel program (Microsoft, Redmont, WA, USA).
4. Discussion
The primary goal of this work was to determine an optimal and importantly, universal decontamination procedure in order to enable the processing of a wide range of different matrices. Standard decontamination methods for the processing samples of clinical material with NaOH and
N-acetyl cysteine could not be used for highly contaminated samples due to the presence of certain environmental micro-organisms (i.e., sporogenic bacteria and moulds). According to previously published data [
56,
57], a combination of different chemical substances would be required.
A common decontamination method using HCl and oxalic acid was also not tested due to previously published worked that indicated it results in low yield of mycobacteria and high proportion of media contamination [
58]. Quaternary ammonium salts were considered and investigated for decontamination, but the following two technical complications were encountered. When decontaminating faecal samples, it was difficult to obtain (1) a homogeneous sample for inoculation and (2) an aliquot for the use of molecular biological method (qPCR). After decontamination with quaternary ammonium salts and final centrifugation, sparingly soluble precipitates often formed, and a large amount of foaming of the processed sample often occurred (unpublished observation), causing the aforementioned issues.
Initially for this study, CPC surfactant was used with 4% NaOH for the decontamination of samples. In these experiments, the decontamination efficiency was not satisfactory because the contamination rate of the culture media exceeded 30% (unpublished data). A “pre-cultivation” method published by Portaels et al. [
59] was also tested. In this method, contaminating microbes are revived and then subsequently eliminated. However, this procedure was unsuccessful, with up to 80% contamination and massive multiplication of other microbes, which made it impossible to isolate mycobacteria (unpublished data). Allen [
60] also noted similarly high sample contamination when decontaminating human faecal samples from patients with human tuberculosis.
Therefore, we developed a new decontamination procedure using a mixture of 4.0% NaOH and 0.5% TDAB. Löwenstein–Jensen media was used for culturing due to its widespread use for culturing mycobacteria. A 4.0% NaOH solution is an established decontamination solution for isolating mycobacteria, and the addition of TDAB enhanced isolation of mycobacteria; however, it was not possible to rule out any possible detrimental (even devitalising) effects that TDAB may have on living mycobacteria present in the examined sample.
Due to the nature and source of the examined samples, an overall contamination rate of 7.4% can be regarded as more than satisfactory. In Cluster R (raw plant material and peat), the low sample contamination rate of only 2.6% is an excellent result. Even the highest contamination rate, seen in samples from Cluster E (animals’ faeces) at 12.6%, is considered acceptable (
Figure 1). In none of the three clusters did the level of contamination exceed mycobacterial growth (
Figure 2).
The degree of media contamination and mycobacterial recovery is comparable to the two-phase method used to treat avian guano with NaOH and cetypyrimidium chloride. Cultivation took place on soils enriched with antibiotics. Cultivation took place on antibiotic-enriched media with contamination reaching 19% and mycobacteria isolation reaching 27.8% [
61]. In our study, we found a similar culture positivity, which reached 29.4% (
Figure 2). In contrast, using our modified decontamination method, we recorded only 12.6% contamination of the samples (
Figure 1).
In contrast to a study by Neumann et al. [
62] comparing decontamination procedures for the treatment of water samples before mycobacterial isolation, a higher proportion of contamination was observed on media cultured at 30 °C. Moreover, in that study, mycobacterial growth was highest at 37 °C. On the other hand, in our study, both slow-growing and fast-growing mycobacteria were isolated especially in the mesophilic temperature range, i.e., at 30 °C (
Table 1).
The differences in the spectrum of captured species of mycobacteria within individual clusters, particularly the higher level of detection in cluster R (peat and plant material), can be explained by known ecological characteristics [
42]. Mycobacterial species are more commonly found in these matrices (
Table 1). From an epidemiological point of view, these matrices, in which these mycobacterial species multiply, represent a natural reservoir [
42,
63]. Clusters S and E are likely vectors for the spread of mycobacterial species, or possibly only transient habitats in which they do not multiply (
Table 1). However, it is possible that the multiplication of these mycobacterial species occurs to a limited extent under certain conditions, such as temperatures between 18 and 20 °C, pH below 7.0, and the presence of some organic substances as a source of nutrients [
42,
64,
65].
Microscopic examination was only useful to evaluable samples containing a high concentration of mycobacterial cells. The presence of large amounts of solid residues and acid-fast-stained (i.e., red) artefacts make microscopic examination difficult to assess specifically in terms of objectivity (
Figure 8). Microscopic examination can be replaced by the determination of the presence of mycobacterial DNA by qPCR. For practical reasons, a commercial product with a standardised protocol and a simple design was used in our study for the examination. The design for this
Mycobacterium assay is based on the publication of Radomski et al. [
20]. The results of the qPCR assay indicated a degree of discrepancy in the detection of the 16S RNA target site using a specific primer 110F/I571R with true capture and growth rate of live mycobacteria.
The primary evaluation of sensitivity and specificity of the PMA-qPCR method compared to the culture determination (considered the gold standard) showed values of 92.2% and 49.9%, respectively (
Table 4). The presence of residues of genetic material from dead mycobacterial cells is the most likely cause of false positives seen in the PMA-qPCR method [
42]. It is also not possible to exclude the interaction of primers with DNA segments of several representatives of the genera
Corynebacterium,
Bacillus and
Flavobacterium very abundant in many matrices. This non-specific detection in the listed genera has also been reported by the authors of the first report of the qPCR method for detection of mycobacteria [
20]. Alternatively, the PMA-qPCR method may be detecting the presence of unculturable mycobacterial species. Cut-off value derived from the model was determined to be 1.75 × 10
3 mycobacterial cells, as measured by the qPCR method. After application of this cut-off value on the raw data, the diagnostic sensitivity decreased to 78.8%, but there was a significant increase in specificity to 77.3% (
Table 5).
The work was performed continuously and retrospectively on natural matrices. We therefore accept that the unambiguous determination of the yield could be skewed by the difficulty of determining the true measurement of the presence of viable mycobacteria in the examined samples (missed blank samples). Yield can be affected of course by the decontamination method due to sensitivity of some mycobacterial species to a decontamination agent [
66]. Cultivation may be (and will be in further work) refined by using special cultivation media (e.g., RGM) [
67]. For detection of mycobacteria by standard qPCR in these samples, the specificity was not high, perhaps unsurprisingly, as this method was primarily designed for less microbially loaded samples. However, a more fundamental shift in detection accuracy occurred with the elimination of the DNA from dead cells after treatment of the sample with PMA before the qPCR. The average decrease in detection in the order of copies was −2 LOG, with sensitivity at 100% and specificity at 80% (shown in
Table 6 and
Figure 8), which is comparable with the values determined by the authors of the concept of this method for quantification of mycobacteria by qPCR [
34].