Isolation of Nontuberculous Mycobacteria from the Environment of Ghanian Communities Where Buruli Ulcer Is Endemic

ABSTRACT

This study aimed to isolate nontuberculous mycobacterial species from environmental samples obtained from some selected communities in Ghana. To optimize decontamination, spiked environmental samples were used to evaluate four decontamination solutions and supplemented media, after which the best decontamination solution and media were used for the actual analysis. The isolates obtained were identified on the basis of specific genetic sequences, including heat shock protein 65, IS2404, IS2606, rpoB, and the ketoreductase gene, as needed. Among the methods evaluated, decontamination with 1 M NaOH followed by 5% oxalic acid gave the highest rate of recovery of mycobacteria (50.0%) and the lowest rate of contamination (15.6%). The cultivation medium that supported the highest rate of recovery of mycobacteria was polymyxin B-amphotericin B-nalidixic acid-trimethoprim-azlocillin–supplemented medium (34.4%), followed by isoniazid-supplemented medium (28.1%). Among the 139 samples cultivated in the main analysis, 58 (41.7%) yielded mycobacterial growth, 70 (50.4%) had no growth, and 11 (7.9%) had all inoculated tubes contaminated. A total of 25 different mycobacterial species were identified. Fifteen species (60%) were slowly growing (e.g., Mycobacterium ulcerans, Mycobacterium avium, Mycobacterium mantenii, and Mycobacterium malmoense), and 10 (40%) were rapidly growing (e.g., Mycobacterium chelonae, Mycobacterium fortuitum, and Mycobacterium abscessus). The occurrence of mycobacterial species in the various environmental samples analyzed was as follows: soil, 16 species (43.2%); vegetation, 14 species (38.0%); water, 3 species (8.0%); moss, 2 species (5.4%); snail, 1 species (2.7%); fungi, 1 species (2.7%). This study is the first to report on the isolation of M. ulcerans and other medically relevant nontuberculous mycobacteria from different environmental sources in Ghana.

IMPORTANCE Diseases caused by mycobacterial species other than those that cause tuberculosis and leprosy are increasing. Control is difficult because the current understanding of how the organisms are spread and where they live in the environment is limited, although this information is needed to design preventive measures. Growing these organisms from the environment is also difficult, because the culture medium becomes overgrown with other bacteria that also live in the environment, such as in soil and water. We aimed to improve the methods for growing these organisms from environmental sources, such as soil and water samples, for better understanding of important mycobacterial ecology.

INTRODUCTION

Nontuberculous mycobacteria (NTM), also known as atypical mycobacteria or mycobacteria other than tuberculosis (TB), belong to the same genus as the pathogens that cause TB and leprosy (1). NTM are ubiquitous in the environment (2), constitute the majority of species (169 species) in the genus Mycobacterium (3), and are largely considered nonpathogenic. They are assuming public health importance, however, as they cause opportunistic infections of all body parts in both humans and animals (1, 4). The abilities of NTM to cause disease vary among species (5, 6). While some species have not been shown to cause disease in humans (Mycobacterium flavescens and Mycobacterium moriokaense), some produce disease occasionally and others almost always cause disease (Mycobacterium avium and Mycobacterium kansasii) (7). The reasons for these variations in pathogenicity are not known.

Although NTM can cause several types of infections in immunocompetent individuals, the emergence of the human AIDS pandemic has made NTM more relevant; M. avium complex was identified as a major cause of opportunistic infections in patients infected with human immunodeficiency virus (HIV) (8). Identified risk factors associated with NTM include but are not limited to (i) reduced immunocompetence as a result of HIV, cancer, chemotherapy, or immunosuppression linked to transplantation; (ii) alcoholism; (iii) preexisting lung diseases, including prior tuberculosis and pneumoconiosis; and (iv) smoking (9, 10). Some infections caused by NTM in immunocompetent individuals in recent times include cervical lymphadenitis and pulmonary infections caused by M. kansasii and Mycobacterium intracellulare (11–13), skin infections caused by Mycobacterium marinum, Mycobacterium ulcerans, and Mycobacterium haemophilum (14–16), and nosocomial infections caused by Mycobacterium chelonae and Mycobacterium fortuitum (17).

Infection with NTM has also been shown to affect the efficacy of the Mycobacterium bovis bacillus Calmette-Guerin (BCG) vaccine (9). The BCG vaccine, which is to protect against tuberculosis (TB), offers different levels of protection among adult human populations (18–21). The variations have been partly attributed to increased exposure to NTM species in some geographical regions (20, 21), reducing protection against pulmonary TB in adults (22). The mechanisms for the decreased BCG efficacy are still not fully understood; however, it has been suggested that exposure to NTM results in a broad immune response that is rapidly recruited after BCG vaccination, thereby controlling the multiplication of the vaccine strain, which is needed for the production of Mycobacterium tuberculosis-specific protective responses (20, 23).

The public health importance of NTM calls for an effective strategy for the prevention and control of NTM infections, which is currently lacking (24). This lack is related to a lack of understanding regarding the mode(s) of transmission. Treatment of NTM infections is complicated by the nonspecificity of the drug regimen (25); moreover, NTM are naturally resistant to some of the available antimycobacterials, and no vaccine is currently available for the prevention of NTM infections (26). Therefore, there is a need for a better understanding of the ecology of NTM, to enhance knowledge regarding disease epidemiology and to guide the design of control and prevention strategies. Different modes of transmission, including ingestion, aspiration, and inoculation of the pathogen from the environment, have been hypothesized for diseases caused by NTM, but none has been proven to date (7). Unlike TB, diseases caused by NTM are rarely, if ever, transmitted from person to person (8).

Isolation and characterization of NTM are essential to improve the understanding of pathogen ecology and transmission. Isolation of NTM from the environment is a major challenge because of the complex mixture of other fast-growing bacteria and fungi in environmental samples, which contaminate the culture medium during isolation (27). Thus, removal of the mixed microorganisms in a decontamination step is crucial for the isolation of NTM. Moreover, attributing an infection to NTM in clinical samples is not straightforward, as NTM from the laboratory or sample collection environment can contaminate the culture. Thicker cell walls and unique mycolic acids (13) contribute to the impermeability, hydrophobicity, and slow growth of both slowly and rapidly growing mycobacteria (28). These characteristic features preferentially enhance NTM attachment to surfaces (29) and resistance to disinfectants, antibiotics, acids, and bases (30), which can be used selectively in the decontamination step.

A major NTM species of public health significance in Ghana and West Africa is M. ulcerans, the causative agent of Buruli ulcer (BU). Like other NTM species, its control is hampered in part because of a lack of understanding of the pathogen ecology and mode of transmission. Isolation of the pathogen from different environmental samples from West Africa has not been successful. To the best of our knowledge, only one study has been reported, in which M. ulcerans was successfully isolated from an aquatic insect (31). Therefore, this study was conducted in communities in which BU is endemic, increasing our chances of isolating M. ulcerans from the environment, to isolate and to profile NTM from the environment in Ghana for the first time.

MATERIALS AND METHODS

Ethics statement.

Ethical clearance for the study was obtained from the institutional review board of the Noguchi Memorial Institute for Medical Research (federal assurance no. FWA00001824).

Study sites.

This study was cross-sectional and was conducted between September 2012 and November 2013, in six communities along two major river basins (Densu and Offin) in Ghana (Fig. 1). The procedure that was used for the site selection was detailed previously by Ampah et al. (32). Community designation as endemic or nonendemic was based on the passive case report data of the National Buruli Ulcer Control Programme (NBUCP) and findings from our active case search activities (32, 33). The following communities along the Densu River basin were included: Ntabea (nonendemic) in the East Akim district, upstream of the river; Ashongkrom (endemic) in the Akwapim South district, midstream of the river; and Domesampaman (endemic) in the Ga-West Municipality of the Greater Accra Region, downstream of the river. All of the studied communities along the Offin River basin were designated endemic; the communities included Ntobroso and Achiase in the Atwima district, upstream and midstream of the river, respectively, and Mfantsiman in the Upper Denkyira district, downstream of the river. Extensive seroepidemiological studies (34–36) conducted in these communities indicated high levels of serological evidence of exposure of community members to M. ulcerans 18-kDa small heat shock protein (hsp) (33), irrespective of the disease burden.

FIG 1.

Map of Ghana, showing the Densu and Offin River basins and selected communities. Map created with ArcGIS 10.0 using GPS coordinates from the National Buruli Ulcer Control Programme (NBUCP).

Sample collection.

Prior to sample collection, all of the selected communities were mapped using the Global Positioning System (GPS) coordinates from the NBUCP and ArcGIS 10.0 mapping software. Sites of frequent human activities, such as bodies of water utilized regularly for domestic purposes such as washing and cooking, communal bathing areas utilized by household members, school compounds (particularly playgrounds), agricultural farms, market grounds, community centers, and sources of drinking water, such as boreholes and water from storage tanks in homes, were used as reference points. Each community was divided into grids, and sampling points were randomly selected, using a randomization tool within ArcGIS 10.0, for environmental sample collection.

Environmental samples were collected by using a convenience approach. We walked through the communities and traced the GPS coordinates from the designated reference points, using the generated grid points for environmental sample collection. At each sample location, we collected a soil sample and any other sample within a 1-m reach. Samples that were more than 1 m away from the location point were excluded from the collection. The collection of samples was carried out for 3 hours each sampling time, starting at 8:00 a.m. and ending at 11:00 a.m.

Samples such as soil (approximately 5 g), water (45 ml), and fungi found growing in the soil and on dead and decaying logs were collected separately into sterile, labeled, 50-ml Falcon tubes (BD Biosciences), which were sealed tightly to avoid cross-contamination of samples. Snail samples were kept in labeled sealable plastic bags. Moss and vegetation parts were collected and pressed separately into labeled 50-ml Falcon tubes (BD Biosciences). We collected snails because aquatic snails have been reported to harbor M. ulcerans transiently (37); therefore, we sought to determine whether edible terrestrial snails also could harbor M. ulcerans. All samples were clearly labeled immediately, were kept in a cool pack at 4°C after collection, and were transported to the laboratory within 6 h after collection. Samples were processed on the day of collection and were analyzed by Ziehl-Neelsen staining and culture within 48 h after collection.

Sample processing.

Snail and fungus samples were diced separately with sterile disposable surgical blades (Swann-Morton) and were homogenized using a sterile porcelain mortar and pestle (Cole-Parmer). The contents were then suspended in 10 ml of phosphate-buffered saline (PBS) (Sigma-Aldrich). Fecal samples were processed like snail and fungus samples. Soil samples were prepared using a modified version of the method described by Parashar et al. (38). Approximately 5-g soil samples were suspended in 30 ml of PBS in sterile 50-ml Falcon tubes (BD Biosciences). Samples were shaken vigorously for 1 min and then centrifuged at 600 × g for 5 min at 4°C, to pellet the soil particles. The turbid supernatants (15 ml) were transferred into new sterile 50-ml Falcon tubes (BD Biosciences) and centrifuged at 7,000 × g for 10 min at 4°C, and the pellets were suspended in 10 ml of PBS (Sigma-Aldrich). Biofilms were prepared from moss and vegetation parts using a modified version of the method described by Gryseels et al. (39). Samples were emptied into sterile plastic resealable bags, and 50 ml of PBS (Sigma-Aldrich) was added to each bag. The contents of the bags were vigorously agitated to dislodge the biofilms into solution. The suspensions were poured into sterile 50-ml Falcon tubes (BD Biosciences) and centrifuged at 1,700 × g for 30 min, to sediment all suspended bacteria. The supernatant was decanted and the resulting pellet was suspended in 10 ml of PBS (Sigma-Aldrich). Water samples were vortex-mixed to homogeneity and centrifuged at 1,700 × g for 30 min, to sediment all suspended bacteria (38). The supernatant was decanted and the resulting pellet was suspended in 10 ml of PBS (Sigma-Aldrich). Prepared sample suspensions were kept at 4°C.

Preparation of spiked environmental samples.

A suspension of in-house reference Mycobacterium ulcerans isolate NM209 was subcultured and harvested in mid-log phase, and a suspension in PBS at a concentration of 10⁸ CFU ml⁻¹ was mixed with 1 ml of 10⁷ CFU ml⁻¹ each of an Escherichia coli suspension obtained from a stool sample, Pseudomonas aeruginosa obtained from a wound culture (representing Gram-negative bacteria), and Gram-positive rods (Bacillus sp.) obtained from an open plate culture. Two-milliliter aliquots of the pooled suspension were then prepared and used for isolation of M. ulcerans. Eight environmental samples (4 water and 4 wet soil samples) were spiked with 10⁸ CFU ml⁻¹ of the M. ulcerans pool suspension and decontaminated.

Optimization of decontamination step.

To optimize the decontamination procedure for the isolation of mycobacteria from environmental samples, we evaluated an in-house decontamination method using NaOH-oxalic acid (40), together with three other standard decontamination protocols, i.e., SDS-NaOH (38), NaOH-malachite green-cycloheximide (41), and N-acetyl-l-cysteine–NaOH (42), using the spiked samples.

NaOH-oxalic acid method.

Two volumes of 1 M NaOH were added to the spiked sample suspension, and the mixture was incubated at room temperature for 20 min, with intermittent vortex-mixing. The reaction mixture was then neutralized by the addition of sterile distilled water to the 45-ml mark of the 50-ml Falcon tube. The sample was centrifuged (Eppendorf 5810 R) at 5,000 × g for 30 min and the supernatant was carefully decanted, after which 2 ml of 5% oxalic was added to the pellet and the mixture was incubated at room temperature for 30 min, with intermittent vortex-mixing. The decontamination reaction was finally terminated by neutralization with sterile distilled water.

SDS-NaOH method.

Two milliliters of 3% SDS-4% NaOH was added to 2 ml of the spiked sample suspension, and the mixture was incubated at room temperature for 30 min, with intermittent vortex-mixing. The decontamination reaction was stopped by neutralization of the mixture with sterilized distilled water to the 45-ml mark of the 50-ml Falcon tube. The suspension was then centrifuged (Eppendorf 5810 R) at 5,000 × g for 30 min at 4°C, after which it was decanted.

NaOH-malachite green-cycloheximide method.

For decontamination with 2% NaOH-0.3% malachite green-0.15% cycloheximide, 2.5 ml was added to 2.0 ml of the spiked sample suspension, and the mixture was incubated at room temperature for 30 min, with intermittent vortex-mixing. The decontamination reaction was stopped by neutralization of the mixture with 1 N HCl.

N-Acetyl-l-cysteine–NaOH method.

A volume of 3 ml of N-acetyl-l-cysteine–NaOH solution was added to the spiked sample suspension, and the mixture was incubated at room temperature for 20 min, with intermittent vortex-mixing. The decontamination reaction was terminated by neutralization of the mixture with sterile distilled water to the 45-ml mark of the 50-ml Falcon tube.

Determination of best culture medium for primary isolation.

In order to determine the most effective cultivation medium for the recovery of mycobacterial species, individual decontaminated samples were centrifuged at 5,000 × g for 30 min. The supernatant was decanted, 1 ml of sterile PBS was added, and 0.1 ml of the pellet-PBS suspension was then inoculated, in duplicate, in prepared Lowenstein-Jensen (LJ) medium (43) supplemented with polymyxin B (2,000,000 IU liter⁻¹)-amphotericin B (10 mg liter⁻¹)-nalidixic acid (10 mg liter⁻¹)-trimethoprim (10 mg liter⁻¹)-azlocillin (10 mg liter⁻¹) (PANTA) (BD), isoniazid (INH) (0.2 mg liter⁻¹) (Sigma-Aldrich), or ethambutol (EMB) (2 μg liter⁻¹) (Sigma-Aldrich) or not supplemented (i.e., drug-free [DF] medium). The inoculated culture tubes were incubated at 37°C and observed for macroscopic growth for 6 months, at which time they were discarded. Acid-fast bacilli (AFB) were detected by preparing smears of bacterial colonies on frosted slides (Sigma-Aldrich), which were allowed to dry and were fixed with heat. The smears were then stained using the Ziehl-Neelsen procedure (44) and graded with the International Union against Tuberculosis and Lung Diseases grading system. A slide was graded as negative only after at least 100 fields were read under oil immersion without the detection of AFB and the results were confirmed by a second reader.

Identification of AFB-positive isolates by GenoType Mycobacterium CM/AS assay.

Genomic DNA from heat-killed AFB isolates was extracted. The GenoType Mycobacterium CM/AS assay (Hain Lifesciences) was performed following the manufacturer's protocol (45). The reaction mixture consisted of 2 μl of MgCl₂, 5 μl of 10× buffer, 35 μl of biotinylated primer-nucleotide mixture, 0.2 μl of HotStar Taq (Qiagen, Hilden, Germany), 2.8 μl of nuclease-free water, and 5 μl of DNA, in a total volume of 50 μl. PCR was performed as follows: 95°C for 15 min, 10 cycles of 95°C for 30 s and 58°C for 2 min, 20 cycles of 95°C for 25 s and 53°C for 40 s, and 20 cycles of 70°C for 8 min. Hybridization and detection were performed in a Twincubator (Hain Lifescience GmbH), following the manufacturer's protocol (45).

Isolation of NTM from environmental samples.

Based on findings from the spiked decontamination evaluation, samples were cultivated using the most effective decontamination method, i.e., the in-house 1 M NaOH-5% oxalic acid method, as previously. During the mock experimentation, we observed fungal growth on cultures inoculated on PANTA-supplemented medium. Therefore, we added the antifungal mycobactin J (2 mg liter⁻¹; IDVet) to the PANTA-supplemented LJ medium (yielding PANTA-mycobactin J [PM] medium). Suspensions were then inoculated in duplicate in prepared LJ medium supplemented with INH, PM medium, or DF medium. The inoculated culture tubes were incubated at 32°C and observed for macroscopic growth for 6 months, at which time they were discarded. The choice of 32°C was to enable isolation of M. ulcerans, which is an important NTM in Ghana. In addition, 2 drops of each sample suspension were put on frosted slides (Sigma-Aldrich), allowed to dry, fixed with heat, and stained using the Ziehl-Neelsen procedure. Culture tubes were read daily during the first week, for the isolation of rapid growers, and thereafter were read weekly. For the cultures that showed growth, the week of growth was recorded and the growth was quantified. Tubes that showed positive growth were purified by subculture of individual colonies on drug-free medium for species identification. AFB were detected by preparing smears of bacterial colonies and staining them for AFB.

Identification of mycobacterial isolates.

The isolates confirmed as AFB positive were harvested, killed by heating at 95°C for 30 min, and used for genomic DNA extraction (46). Cell wall lysis was achieved first by the addition of 300 μl of lysis buffer to bacterial pellets, followed by the addition of 50 μl of lysozyme (10 mg ml⁻¹; Fluka), 100 μl of SDS (0.04 mg ml⁻¹), and 40 μl of proteinase K (0.2 mg ml⁻¹; Roche). The mixture was incubated at 37°C for 1 h, after which 200 μl of zirconia beads (Sigma-Aldrich) was added and the mixture was homogenized at 4,700 × g for 4 min, using a FastPrep-24 homogenizer (MP Biomedicals). The mixture was centrifuged at 5,000 × g for 5 min, and the supernatant, containing released DNA, was transferred to a new Eppendorf tube. Proteins and other cellular debris were removed by two steps of phenol-alcohol extraction, with the addition of 500 μl of phenol-chloroform-isoamyl alcohol (Sigma) to the mixture and centrifugation at 5,000 × g for 10 min. DNA precipitation was achieved with the addition of 20 μl of 3 M sodium acetate (Sigma) to the mixture, followed by the addition of 500 μl of absolute ethanol (Sigma-Aldrich). The mixture was frozen at −20°C overnight and centrifuged at 5,000 × g for 30 min, after which the pellets were dried at 50°C. The genomic DNA was suspended in 100 μl of nuclease-free water. The concentrations of the DNA were checked using a NanoDrop 2000C spectrophotometer (Thermo Scientific).

A 441-bp portion of mycobacterial heat shock protein 65 (hsp65) was amplified using the Mycobacterium genus-specific TB-11 and TB-12 forward and reverse primers (Table 1) (47), as described previously (48). The PCR mixture contained 5 μl of a 1:100 dilution of template DNA, 0.25 μM concentrations of each primer, 6 μl of Q-solution, 3 μl of 10× buffer, 200 μM (each) dATP, dCTP, dGTP, and dTTP (Pharmacia Biotech), 1.5 mM MgCl₂, and 0.5 U of Fire Taq polymerase, in a total volume of 100 μl. Amplification was performed using 32 cycles of 5 min at 94°C, 30 s at 94°C, 30 s at 60°C, 1 min at 72°C, and 10 min at 72°C, in an Applied Biosystems 2720 thermal cycler. Amplified product (10 μl) was confirmed by gel electrophoresis. The amplified PCR product (40 μl) was sequenced by outsourcing. The generated sequences were edited using CodonCode Aligner 6.0.2 software to remove vector sequences, and species were identified by using NCBI Microbial Nucleotide BLAST, using default settings (49).

TABLE 1.

Primers used for NTM detection and identification

Primer no.	Primer name	Sequence (5′ to 3′)	Reference
1	TB-11F	ACCAACGATGGTGTGTCCAT	46
2	TB-12R	CTTGTCGAACCGCATACCCT
3	IS2404-F	GGCAGTTACTTCACTGCACA	49
4	IS2404-R	CGGTGATCAAGCGTTCACGA
5	IS2606-F	GGTGCGGTTCCATTGAGA	46
6	IS2606-R	GTCGTAGATGTGGGCGAAAT
7	KR-F	TTCTGTTCCCGCAGTTCTTT	This work
8	KR-R	CTAGCTCGGAAGTGGTCAGC
9	rpoB-F	GTGGGTCGGTACAAGGTCAA	This work
10	rpoB-R	GCGGTCAGGTAGTGGATCTC

Confirmation of M. ulcerans with IS2404, IS2606, ketoreductase B domain, and rpoB.

The M. ulcerans isolate obtained was confirmed using PCR detecting the following gene targets: IS2404 (50), IS2606 (47), ketoreductase (KR) B domain, and rpoB. The primers for KR and rpoB were designed using Primer3 software (50). The primer sets used for the amplification are shown in Table 1. The reaction mixtures for all of the targets (IS2404, IS2606, KR, and rpoB) contained 10 μM concentrations of each primer, 200 μM (each) dATP, dCTP, dGTP, and dTTP, 19 μl of nuclease-free water, 25 mM MgCl₂, 3 μl of 10× buffer, 1 U Fire PolTaq (Solis BioDyne), and 1 μl of template DNA (5 ng), in a total volume of 30 μl. Amplification was performed in an Applied Biosystems 2720 thermal cycler, as follows: 32 cycles of 5 min at 94°C, 30 s at 94°C, 30 s at 60°C, 1 min at 72°C, and 10 min at 72°C. Amplified DNA (10 μl) was subjected to electrophoresis in a 2% agarose gel prepared with Tris-borate-EDTA (TBE) buffer and was detected by ethidium bromide staining and UV transillumination.

Data analysis.

The sensitivities of the four decontamination methods and cultivation media were evaluated. The colony morphology and the length of time before visible colonies appeared were recorded during culture readings. Cultures were said to be positive if at least one tube from the same treatment confirmed AFB growth, cultures were said to be negative if none of the culture tubes had microbial growth after 6 months, and cultures were classified as contaminated when all of the tubes had more than one-half of the tube with other bacterial growth or liquefied contents. The performance of decontamination methods and cultivation media was assessed using the t test.

RESULTS

Performance of decontamination methods and growth media tested with spiked samples.

Using the four decontamination procedures and four different culture media, a total of 128 cultures were performed for the 8 samples analyzed. As indicated in Table 2, the decontamination procedure that resulted in the smallest number of contaminated tubes was the NaOH-oxalic acid method (5 cultures [15.6%]); the method that gave the highest no-growth rate was the malachite green-cycloheximide method (11 cultures [34.4%]), and the SDS-NaOH method gave the largest number of contaminated tubes (28 cultures [87.5%]). The cultivation medium that supported the highest rate of mycobacterial growth was PANTA-supplemented medium, followed by INH-supplemented medium, drug-free medium, and EMB-supplemented medium. AFB isolates from 2 of the 8 samples were confirmed as M. ulcerans; in both cases, the medium was PANTA-supplemented medium.

TABLE 2.

Decontamination methods and cultivation media for recovery of Mycobacterium ulcerans

Decontamination methoda	No. of culturesb
	Positive					Contaminated					No mycobacterial growth
	DF	INH	EMB	PANTA	Total (n = 32)	DF	INH	EMB	PANTA	Total (n = 32)	DF	INH	EMB	PANTA	Total (n = 32)
SDS-NaOH	0	0	0	0	0	6	8	8	6	28	2	0	0	2	4
NaOH-Mal-C	2	0	0	5	7	4	1	5	1	11	2	2	3	7	14
NaOH-OA	3	5	3	5	16	1	1	3	0	5	4	2	2	3	11
NALC-NaOH	2	4	2	1	9	3	4	3	2	12	3	3	3	2	11

^aSDS-NaOH, sodium dodecyl sulfate-sodium hydroxide; NaOH-Mal-C, NaOH-malachite green-cycloheximide; NaOH-OA, NaOH-oxalic acid; NALC-NaOH, N-acetyl-l-cysteine–sodium hydroxide. The NaOH-oxalic acid method was the most effective among the evaluated procedures.

^bDF, drug-free; INH, isoniazid; EMB, ethambutol. The cultivation medium with the highest mycobacterial growth rate was PANTA-supplemented medium, followed by INH-supplemented medium, and finally DF medium. EMB-supplemented medium had the lowest number of positive cultures and the highest number of contaminated cultures.

Direct smear microscopy.

Direct smear analysis of the 139 samples that were cultivated identified only 12 (8.6%) as containing AFB. Acid-fast bacilli were detected mainly among moss, snail, soil, and vegetation samples, as indicated in Table 3. Slides of some samples in which AFB were detected are shown in Fig. 2.

TABLE 3.

Rates of positivity for acid-fast bacilli by direct smear analysis and isolation of mycobacterial species

Sample type	Total no. of samples	No. (%) of samples positive for acid-fast bacilli/total no.	No. of samples yielding mycobacterial isolate/total no.
Fungi	4	0/4 (0)	1/4
Snails	5	1/5 (20)	1/5
Moss	8	1/8 (12.5)	3/8
Soil	54	6/54 (11.1)	30/54
Vegetation	56	4/56 (7.1)	21/56
Water	12	0/12 (0)	2/13
Total	139	12/139 (8.6)	58

FIG 2.

Direct smear analyses, using the Ziehl-Neelsen procedure, of environmental samples from water (A), moss (B), vegetation (C), and snail (D). Red arrows, acid-fast bacilli. Magnification, ×1,000.

Isolation of Mycobacterium spp. from environmental samples.

Of the 139 environmental samples cultivated, 58 (41.7%) yielded mycobacterial growth, 70 (50.4%) had no bacterial growth, and 11 (7.9%) had all of the inoculated tubes contaminated. The performance of the different growth media, as indicated in Table 4, showed that the cultivation medium supporting the greatest recovery of mycobacterial growth was LJ medium supplemented with PANTA and mycobactin J (i.e., PM medium) (75 samples [26.9%]), followed by INH-supplemented medium (63 samples [22.7%]) and finally DF medium (25 samples [8.9%]). The smallest number of contaminated tubes was recorded with the mycobactin J and PANTA supplement (111 tubes [39.9%]), followed by INH-supplemented medium (132 tubes [47.5%]) and DF medium, giving the greatest number of contaminated tubes (224 tubes [80.6%]) (P < 0.001).

TABLE 4.

Performance of different growth media in isolating Mycobacterium spp. from the environment

Mediuma	No. of culture tubes
	Positive	Contaminated	No growth	Total
DF	24	224	30	278
INH	63	132	83	278
PM	75	111	92	278
Total	162	467	205	834

^aDF, drug-free; INH, isoniazid; PM, PANTA-mycobactin J. The cultivation medium with the highest mycobacterial growth rate was PM medium, followed by INH-supplemented medium and finally DF medium.

Isolated mycobacterial species.

Based on colonial morphology, growth rate characteristics, and growth medium findings, a total of 162 AFB-positive isolates (obtained from 58 samples) were collected for identification. Analysis of hsp65 sequences identified 25 different mycobacterial species, as shown in Fig. 3. The mycobacterial species isolated were from soil (16/30 isolates), vegetation (14/21 isolates), water (3/2 isolates), moss (2/3 isolates), molluscs (1/1 isolate), and fungi (1/1 isolate). The isolates consisted of slowly growing mycobacteria (15 isolates [60%]) and rapidly growing mycobacteria (10 isolates [40%]), with the greatest recovery being mycobacteria found among soil and vegetation samples, as shown Fig. 3. Two M. ulcerans isolates were isolated from moss and soil samples from communities in which BU is endemic. Other slowly growing mycobacteria, including M. avium, Mycobacterium mantenii, and Mycobacterium malmoense, and rapidly growing mycobacteria such as M. chelonae, Mycobacterium abscessus, and M. fortuitum were among the many species isolated (Table 5).

FIG 3.

Heat shock protein 65 analysis, revealing 25 different mycobacterial species isolated from environmental samples from communities in which BU is endemic. *, causes skin and soft tissue infections.

TABLE 5.

Growth characteristics of and isolation media for Mycobacterium spp.

Mycobacterium sp.	Isolation medium(a)a
Slowly growing
M. ulceransb	PM
M. szulgaib	PM
M. mantenii	PM
M. senuense	PM
M. lentiflavumb	PM, DF
M. aviumb	PM, DF, INH
M. engbaekii	PM, DF
M. gordonae	PM, INH, DF
M. paraense	PM
M. terraeb	DF
M. asiaticus	DF, INH
M. nonchromogenicumb	DF
M. intracellulare	INH
M. genavense	INH
M. malmoense	PM
Rapidly growing
M. peregrinum	PM
M. chelonaeb	PM, DF, INH
M. abscessusb	PM, DF, INH
M. fortuitumb	DF, INH
M. smegmatis	DF, INH
M. septicum	PM, INH
M. porcinum	INH
M. setenseb	DF, INH
M. boenickeib	PM, INH
M. arupense	PM

^aPM, PANTA-mycobactin J; DF, drug-free; INH, isoniazid.

^bA mycobacterial species that causes skin and soft tissue infections.

Among the media tested, LJ medium supplemented with both PANTA and mycobactin J (PM medium) gave the highest rates of recovery of mycobacterial species (16 samples [42.1%]), followed by INH-supplemented medium (13 samples [34.2%]) and drug-free LJ medium, with the lowest rate (11 samples [28.9%]). While some mycobacterial species (M. ulcerans, M. mantenii, Mycobacterium szulgai, Mycobacterium senuense, Mycobacterium terrae, and Mycobacterium genavense) were isolated from a single medium, other species (M. chelonae, M. abscessus, M. fortuitum, M. avium, Mycobacterium gordonae, and Mycobacterium septicum) were isolated from the different cultivation medium, as shown in Table 5. Eight (32%) of the species were isolated from PM medium only, 3 (12%) from INH-supplemented medium only, and 2 (8%) from DF medium only, as indicated in Table 6. While 4 mycobacterial species (16%) were isolated from all three cultivation media (PM medium, INH-supplemented medium, and DF medium), 2 (8%) were isolated from PM medium and DF medium or PM medium and INH-supplemented medium and 4 (16%) were isolated from DF medium and INH-supplemented medium (Table 6).

TABLE 6.

Cultivation media and isolated mycobacterial species

Cultivation medium(a)a	No. (%) of mycobacterial species isolated
PM	8 (32)
INH	3 (12)
DF	2 (8)
PM, DF	2 (8)
PM, INH	2 (8)
DF, INH	4 (16)
PM, INH, DF	4 (16)

^aPM, PANTA-mycobactin J; DF, drug-free; INH, isoniazid.

Confirmation of M. ulcerans.

Gel-based PCR analyses of the M. ulcerans isolates from moss sample using primers for the targets IS2404, IS2606, KR, and rpoB, which are found in all mycobacteria, showed products with their respective band sizes on electrophoresis. Product sizes of 492 bp, 332 bp, 615 bp, and 606 bp were detected for IS2404, IS2606, KR, and rpoB, respectively, as shown in Fig. 4.

FIG 4.

Confirmation of M. ulcerans by PCR detection and amplification of IS2404, IS2606, ketoreductase B domain, and rpoB. Lane M, size ladder (Gibco); lane A, M. avium isolated from a cocoa pod; lane B, M. gordonae isolated from soil ; lane C, M. ulcerans isolated from moss; lane D, M. chelonae isolated from soil.

DISCUSSION

The objective of the study was to optimize the procedure for isolation of NTM from the environment by evaluating different decontamination solutions and antibiotic-supplemented media. A secondary objective was to use the best of the evaluated methods to isolate and to profile NTM in environmental samples from communities in Ghana in which BU is endemic. The main achievement of this work was the isolation of different NTM species, including M. ulcerans, from various environmental sources in Ghana for the first time.

Buruli ulcer infections caused by M. ulcerans have afflicted many individuals, particularly in countries of West Africa, including Ghana, in which BU is endemic (15). Over the years, a number of studies aiming to elucidate the ecology and mode(s) of transmission of M. ulcerans have proved futile. Theories that have been propounded to explain the mechanism of M. ulcerans transmission include (i) inhalation or ingestion of aerosolized M. ulcerans from contaminated water (51), (ii) acquisition of M. ulcerans through an insect or animal bite (52), and (iii) contamination of existing wounds or sites of trauma by environmental factors, such as soil, vegetation, and water (53); such theories have not been proved experimentally. One challenge has been the presence of fast-growing microorganisms that attain macroscopic growth within 18 to 48 h, thereby contaminating cultures, inasmuch as the mycobacterial species of interest (M. ulcerans) grows very slowly, requiring more than 6 months to achieve macroscopic growth in some instances. Our isolation of M. ulcerans from moss and soil samples from an area in which BU is endemic supports the concept that M. ulcerans transmission could possibly be through contamination of an open cut or wound by environmental factors, especially among children playing on the ground and agricultural farmers who are in constant contact with such factors. This achievement is a major milestone for further studies to increase the number of isolates, which will allow molecular epidemiological studies aimed toward an understanding of transmission dynamics.

Decontamination using 1 M NaOH followed by a simplified oxalic acid (40) method allowed us to isolate M. ulcerans from both spiked and nonspiked samples. One important factor that is considered in the isolation of mycobacteria from the environment is the decontamination technique employed. Members of the genus Mycobacterium have thick cell walls that are rich in lipids, which allows them to be impermeable to many acids, bases, and detergents (14); this has made decontamination with these agents a common practice in primary cultivation of mycobacteria, to reduce contamination (38, 54). As a result, samples have been decontaminated by different procedures and have been evaluated in different laboratories for decades, with different results (38, 54, 55). Various decontaminating agents, due to their harshness, destroy a substantial number of mycobacteria along with the contaminants, while others are too weak to eliminate contaminants. A good decontamination agent is one that effectively removes unwanted microorganisms and maximizes the recovery of wanted mycobacteria. Decontamination of 139 PCR-positive samples with 1 M NaOH followed by 5% oxalic acid yielded 58 samples (41.7%) that were positive for mycobacteria and 70 (50.4%) with no growth, with only 11 samples (7.9%) being contaminated. Although this procedure was the best among the four different decontamination procedures evaluated (Table 2), the proportion of samples with no growth indicates that it is also very harsh. In a study by Portaels et al., the use of the NaOH decontamination method and the addition of mycobactin to the culture medium did not improve the recovery of mycobacteria from soil samples (55). In our study, however, the addition of the antifungal mycobactin J and the antibiotic PANTA to the cultivation medium was effective in limiting contamination by fungi and fast-growing bacteria, compared to drug-free medium, as shown in Table 4. Isoniazid (INH) was incorporated into the selective medium because some studies have shown that some NTM (for example, M. ulcerans) are resistant to this antibiotic (55, 56) and it could inhibiting nontarget microorganisms; this was found to be true in our study, as INH-supplemented medium performed better than DF medium. Therefore, we recommend the use of our in-house decontamination procedure and LJ medium supplemented with PANTA and mycobactin J to enhance the recovery of NTM, particularly M. ulcerans, from environmental samples.

Species identification of the acid-fast isolates revealed that at least one NTM species was isolated in each type of environmental sample. This finding further implicates the environment as the natural reservoir of NTM, as has been reported from other studies (57, 58). We isolated the greatest number of mycobacterial species from soil (16 species [43.2%]), as indicated in Fig. 3. This is not surprising, as the hydrophobic nature of NTM preferentially confers on them the ability to attach to soil surfaces in order to thrive in the environment (3). Also, the high rate of NTM recovery from soil could be partly due to the abundance of organic matter in the soil, which serves as a nutrient source for the mycobacteria in the environment, since most NTM are known to be saprophytic in nature (38, 59). The high rates of the presence of NTM, especially potentially pathogenic species, in the soil are of public health concern. Agricultural farmers, miners in the pursuit of their work, and children, particularly during playtime, are constantly exposed to soil and vegetation in the environment. Therefore, it is important for persons with frequent body contact with such environmental sources to wear protective clothing, which has been shown to offer some form of protection against Buruli ulcer disease (60). In addition to M. ulcerans, we isolated a number of other NTM species that are known to cause disease in humans, including M. fortuitum, M. chelonae, M. abscessus, M. malmoense, and M. avium. A study by Kankya et al. reported on the isolation of similar NTM species of public health significance in Uganda, although M. ulcerans was not isolated in that study (61).

Direct microscopic analysis of the 139 samples identified only 12 (8.6%) as containing acid-fast bacilli (AFB), although all were confirmed by PCR analysis to contain mycobacteria. The low sensitivity of Ziehl-Neelsen microscopy for detection of AFB is well documented, i.e., samples are positive only when they contain at least 10⁴ CFU ml⁻¹ AFB. This finding suggests that, while NTM species may be abundant in the environment, the bacterial concentrations are quite low. This finding may not be unique to the isolated NTM, as several studies conducted in Ghana and other African countries in which BU is endemic detected low cycle threshold (C_T) values for M. ulcerans, the causative agent of Buruli ulcer, in the environment (62). In contrast, however, studies from Australia reported high C_T values for M. ulcerans in the environment (63).

In conclusion, we isolated M. ulcerans and other slow-growing and fast-growing medically relevant NTM from BU-burdened communities in Ghana. The in-house decontamination procedure (1 M NaOH followed by 5% oxalic acid) and cultivation media supplemented with various antimicrobial and antifungal agents were effective for the recovery of NTM. The isolation of at least one NTM species from each pool of environmental samples cultivated confirms the abundant distribution of NTM in the environment, but their concentrations in the environment may be low, as shown by direct smear analysis.