Assessment of antimycobacterial activities of pure compounds extracted from Thai medicinal plants against clarithromycin-resistant Mycobacterium abscessus

Pure compounds extracted from local medicinal plants

A total of 23 pure compounds, 22 isolated from four medicinal plants (A. monophylla, P. filamentosa, A. conyzoides and R. wittii) and one compound isolated from the cultured mycelium of a luminescent mushroom (N. nambi), were evaluated (Table 1 and Fig. S1). The compounds were purified using column chromatography techniques to isolate the pure secondary metabolites, as described previously (Sombatsri et al., 2018). Stock solutions were prepared by dissolving pure compounds in dimethyl sulfoxide (DMSO), which can dissolve both polar and non-polar compounds.

Bacterial isolates

Mycobacterium abscessus clinical isolates were retrieved from the culture collection of Srinagarind Hospital, Khon Kaen University, Thailand. The colony morphology of each isolate was noted. The identification of M. abscessus species was performed according to the protocol previously published (Kham-Ngam et al., 2019). All isolates were sub-cultured on Lowenstein-Jensen (LJ) media and incubated at 37 °C for 7 days before further analysis. Informed consent was not required for this study. Anonymized, left-over specimens were used. All specimens, including isolates and blood samples, had been obtained during routine practice. This study was approved by the Khon Kaen University Ethics Committee for Human Research (HE611496).

Antibiotic susceptibility testing

The minimum inhibitory concentration (MIC) for clarithromycin was determined according to published protocols (Kham-Ngam et al., 2019). The broth-microdilution method using a RAPMYCOI Sensititre 96-well plate (TREK Diagnostic Systems, Independence, OH, USA) following the manufacturer’s protocol. Briefly, individual colonies of M. abscessus were suspended in demineralized water to obtain a density corresponding to McFarland Standard 0.5. Then, 50 μL of cell suspension were transferred into a tube of cation-adjusted Mueller-Hinton broth (TREK Diagnostic Systems, Independence, OH, USA) with TES buffer (to optimize the conditions for antibiotic stability) to achieve a final cell concentration of approximately 5 × 10⁵ CFU/mL. One-hundred microliters of this inoculum were added to each well of a 96-well plate containing different concentrations of antibiotics. The 96-well microtiter plates were then incubated at 37 °C for 3–14 days under aerobic condition. Clarithromycin susceptibility was read at 3, 5, and 14 days according to CLSI guidelines (CLSI, 2018). A reading at day 3 was used to test for inducible resistance according to previously described protocols (Ananta et al., 2018). Inducible resistance was inferred by changes in MIC values from “susceptible” at day 3 to “resistant” at day 14. Isolates that were resistant on day 3 and thereafter were regarded as demonstrating acquired resistance. All clarithromycin-susceptible and clarithromycin-resistant (both inducible and acquired resistance) M. abscessus clinical isolates were used for further analysis.

Antimycobacterial assay

The antimycobacterial assay was carried out using a broth-microdilution method to determine the MIC values according to the CLSI guidelines (CLSI, 2018). Two-fold serial dilutions of pure compounds were prepared directly in a 96-well microtiter plate. For preparation of the M. abscessus inoculum, the same protocol as described above was used. One-hundred microliters of this inoculum were mixed with 100 μL of pure compound (to give the final concentrations: 1, 2, 4, 8, 16, 32, 64 and 128 μg/mL) and were then added to each well of the 96-well plate. Following incubation for 3–5 days at 37 °C, MICs were visually determined as the lowest concentration of the compound that completely inhibited the mycobacterial growth (Ananta et al., 2018).

Hemolytic assay

Hemolytic activity was determined according to protocols published previously with some modification (Lima Viana et al., 2018). Briefly, 6 mL of blood from a single healthy volunteer were collected and transferred into a heparin collection tube. Whole blood was centrifuged at 5,000 rpm for 5 min and the plasma was then discarded. Concentrated red blood cells (RBCs) were isolated and washed three times with 1% sterile phosphate buffer saline (PBS) solution (pH 7.4) and centrifugation. Then, the RBCs were diluted in 1% PBS to a 5% final concentration of RBC suspension for analysis (950 μL of 1% PBS and 50 μL of concentrated RBCs in a total volume of 1 mL). Two hundred microliters of this RBC suspension were transferred into the tubes containing different concentrations of pure compounds in a total volume of 1 mL (at concentrations before adding the RBC suspension: 1, 2, 4, 8, 16, 32, 64 and 128 μg/mL in 1% PBS). Positive and negative controls were used, these being 1% Triton X-100 solution (Lima Viana et al., 2018) and 1% PBS, respectively. The final volume of each experiment was 1.0 mL. The solutions were incubated at 37 °C for 1 h. After incubation, the suspensions were then centrifuged at 3,000 × g for 2 min. Then, 100 μL of supernatant from each tube were transferred into a 96-well plate for measurement of the absorbance at 540 nm using a microplate reader (each absorbance was measured twice). In addition, RBC morphology was observed under a light microscope and recorded. All tests were performed in duplicate for each test compound. Hemolytic activity was calculated by the following equation:

$$\text{Hemolysis\hspace{0.17em}}(\%)=[(As-An)/(Ac-An)]\times 100$$

where As refers to the absorbance of the sample, An refers to the absorbance of the negative control (RBCs with PBS) and Ac refers to the absorbance of the positive control (RBCs with Triton X-100).

Cell viability assay

To assay the toxicity of each tested compound for human white blood cells, the trypan-blue exclusion test was used (Strober, 2015). Briefly, white blood cells (WBCs) were isolated from 6 mL of blood from a healthy volunteer using the Ficoll density-gradient technique (Boyum, 1976). Whole blood was carefully transferred into a tube containing Ficoll solution (ratio 1:1). Then, the cells were centrifuged at 1,500 rpm for 10 min at 20 °C and the WBC layer was transferred into a new tube. The concentrated WBCs were washed three times with 1% PBS at 1,500 rpm for 5 min at 20 °C and re-suspended in 1 mL of RPMI-1640 media (Gibco™, New York, USA). Fifty microliters of WBC suspension were transferred into individual wells of a 96-well plate and then 50 μL of pure compounds at different concentrations (ranging from 1 to 128 μg/mL) were added. The 96-well plate was incubated at 37 °C for 1 h. Then, 20 μL of the suspension was mixed with 20 μL of 0.4% trypan blue solution in buffered isotonic salt solution (0.81% NaCl and 0.06% K₂HPO₄) and incubated for 3 min at room temperature. Viable and dead cells were counted under a light microscope using a hemocytometer. As a negative control, WBC suspension was treated with PBS. The test was performed in duplicate for each test compounds. Dead cells were calculated using the following equation:

$$\text{Dead\hspace{0.17em}cells\hspace{0.17em}}(\%)\text{=}(\text{number\hspace{0.17em}of\hspace{0.17em}dead\hspace{0.17em}cells/total\hspace{0.17em}number\hspace{0.17em}of\hspace{0.17em}cells})\times \text{100}$$

Genome sequencing and analysis

Genomic DNA of the 30 M. abscessus clinical isolates was extracted using the cetyltrimethylammonium bromide-sodium chloride (CTAB) method (De Almeida et al., 2013) and was sent for genome sequencing (NovogeneAIT, Hong Kong) using an Illumina HiSeq platform generating 150-bp paired-end reads.

The quality of raw sequences was checked using FastQC version 0.11.7 (Andrews, 2010). Trimmomatic (v0.36) software (Bolger, Lohse & Usadel, 2014) was used to remove low-quality reads. High-quality paired-end reads were then mapped to the M. abscessus ATCC 19977 reference genome (GenBank accession number CU458896.1) using BWA-mem (v.0.7.17) (Li, 2013). For converting SAM to BAM format, sorting and indexing the bam files, SAMtools v0.1.19 algorithm was used (Li et al., 2009). GATK version 4.0.5. (McKenna et al., 2010) was used for realignment, generating coverage statistics and mapping details. Both GATK and SAMtools were used for variant calling and filtering, including single-nucleotide polymorphisms (SNPs) and small indels (Li et al., 2009; McKenna et al., 2010).

For phylogenetic analysis, a WGS-based phylogeny was analyzed using mpileup, VCF and coverage files. Maximum-likelihood analysis was performed using MEGA-7 (Qasim et al., 2018) with the general time-reversible (GTR) and gamma model. Support for individual nodes was assessed using 1,000 bootstrap replicates. The phylogenetic tree was visualized using iTol software (https://itol.embl.de/).

Synergism

Combinations of plant secondary metabolites and clarithromycin were evaluated using a microdilution checkerboard method (Garcia, 2010). Five concentrations of each test compound (ranging from 32 to 512 μg/mL) and eight concentrations of clarithromycin (Sigma-Aldrich, St. Louis, MO, USA) (ranging from 32 to 4,096 μg/mL) were prepared, and M. abscessus cell suspensions, prepared as for previous experiments, were used. Fifty microliters of each pure compound and clarithromycin were mixed in a 96-well plate, and 100 μL of inoculum were then added (final concentration of each pure compound ranged from 8 to 128 μg/mL and of clarithromycin ranged from 4 to 1,024 μg/mL). The plate was incubated at 37 °C for 7 days (for isolates with acquired resistance) or 14 days (for isolates with inducible resistance). The results were recorded and interpreted as the fractional inhibitory concentration index (FICI) (Doern, 2014). The FICI value was calculated using the following equation:

$\mathrm{F}\mathrm{I}\mathrm{C}\mathrm{I}=[\mathrm{A}]/(\mathrm{A})+[\mathrm{B}]/(\mathrm{B})$

where [A] refers to MIC (A) in combination with (B), (A) refers to MIC (A) alone, [B] refers to MIC (B) in combination with (A), and (B) refers to MIC (B) alone.

FICI values of ≤0.5, >0.5–1.0, >1.0–4.0 and >4.0 were interpreted as indicating “synergy”, “additive”, “indifference” and “antagonism”, respectively.

Data analysis

All quantitative data are reported as means. Comparison of quantitative data among groups of the toxicity assays was performed using one-way ANOVA followed by post-hoc LSD tests. Number of isolates with colony morphotypes and the susceptibility to the compounds and clarithromycin were compared using Fisher’s exact test. P-values <0.05 were considered statistically significant. All statistical analyses were performed using SPSS version 19.0 (IBM, Armonk, NY, USA).

Identification and characteristics of M. abscessus isolates used

Thirty clinical isolates of M. abscessus on LJ solid medium were recovered. Characteristics of these isolates are described in Fig. 1. Thirteen isolates were identified as belonging to M. abscessus subsp. abscessus and 17 isolates to M. abscessus subsp. massiliense (Fig. 1). Based on their phylogenetic relationships, these isolates were not clonal strains (Fig. 1). The MIC values for clarithromycin among the M. abscessus isolates ranged from 0.12 to ≥16 μg/mL. Ten isolates were phenotypically susceptible to clarithromycin (ranging from 0.12 to 2 μg/mL), while eleven and nine isolates exhibited inducible and acquired resistance to clarithromycin (both ≥16 μg/mL), respectively (Fig. 1 and Table 2).

Antimycobacterial activity of plant secondary metabolites

The MIC values of 23 pure compounds were determined by broth microdilution against 30 clinical isolates of M. abscessus, including clarithromycin-susceptible and-resistant isolates. The MIC cut-off value of 128 µg/mL for at least one isolate was chosen for the selection of potentially useful test compounds. Of the 23 pure compounds, only five (RL008, RL009, RL011, RL012 and RL013) had MIC values lower than 128 µg/mL in any tested isolates. These five compounds were particularly effective in suppressing M. abscessus. The MIC values for these compounds ranged from <1 to >128 µg/mL (Table 2). The five effective pure compounds were selected for further analysis. No significant association was observed between colony morphotypes of M. abcessus isolates and their susceptibility to the test compounds and clarithromycin (P-value >0.05, Table S1).

Toxicity testing on human RBCs and WBCs

The hemolytic activity of the five selected compounds (RL008, RL009, RL011, RL012 and RL013) on human RBCs was evaluated. None showed any hemolytic effect on RBCs (<1%) at the various concentrations used (at the compound concentration before adding RBC suspension: 1, 2, 4, 8, 16, 32, 64, and 128 µg/mL) (Fig. 2 and Table S2). Indeed, the percentage of hemolysis was less than in the negative controls (RBC suspension in 1% PBS that was used as the baseline). RBC morphologies under the light microscope were displayed as the cell shrinks (Fig. 3). These findings indicate that the compounds were not harmful to human RBCs.

For WBCs, the percentage of dead cells following exposure to each tested compound is presented in Fig. 4 and Table S3, and the statistical analysis together with ANOVA with LSD post-hoc multiple comparisons revealed significantly differed among compounds (Fig. 4F). The toxic effects were concentration-dependent. All five compounds caused death of 15–20% of cells at the final concentrations of 64–128 μg/mL, except for compound RL013, for which the mortality rate was significantly lower (3–8%). These results suggest that the compounds have a low toxicity towards human WBCs.

Synergistic antimycobacterial activity of pure compounds and clarithromycin

Based on phenotypic results, fourteen M. abscessus isolates (seven isolates with inducible and seven isolates with acquired resistance to clarithromycin) (Table 2) were randomly selected to study the synergistic effect against these of five pure compounds in combination with clarithromycin. Results (Table 3) showed that the highest degree of synergism was observed for the RL008/CLA and RL009/CLA combinations (FICI ranging from 0.13 to 0.50), which inhibited eight M. abscessus isolates (57%) (Fig. 1). The second strongest synergistic activity was observed for the RL012/CLA combination, followed by the RL013/CLA combination, which showed synergistic effects against seven (50%) and six isolates (42.9%), respectively. The RL011/CLA combination showed the lowest synergistic effect, inhibiting only five isolates (35.7%).