Mutations in Genes Encoding 23S rRNA and FadD32 May be Associated with Linezolid Resistance in Mycobacteroides abscessus

Abstract

Linezolid is one of the antibiotics used to treat the Mycobacteroides abscessus infection. However, linezolid-resistance mechanisms of this organism are not well understood. The objective of this study was to identify possible linezolid-resistance determinants in M. abscessus through characterization of step-wise mutants selected from a linezolid-susceptible strain, M61 (minimum inhibitory concentration [MIC]: 0.25 mg/L). Whole-genome sequencing and subsequent PCR verification of the resistant second-step mutant, A2a(1) (MIC: >256 mg/L), revealed three mutations in its genome, two of which were found in the 23S rDNA (g2244t and g2788t) and another one was found in a gene encoding the fatty-acid-CoA ligase FadD32 (c880t→H294Y). The 23S rRNA is the molecular target of linezolid and mutations in this gene are likely to contribute to resistance. Furthermore, PCR analysis revealed that the c880t mutation in the fadD32 gene first appeared in the first-step mutant, A2 (MIC: 1 mg/L). Complementation of the wild-type M61 with the pMV261 plasmid carrying the mutant fadD32 gene caused the previously sensitive M61 to develop a reduced susceptibility to linezolid (MIC: 1 mg/L). The findings of this study uncovered hitherto undescribed mechanisms of linezolid resistance in M. abscessus that may be useful for the development of novel anti-infective agents against this multidrug-resistant pathogen.

Abbreviations

CLSI: Clinical and Laboratory Standards Institute; MIC: minimum inhibitory concentration; rRNA or rDNA: ribosomal RNA or ribosomal DNA

Introduction

Mycobacteroides abscessus (a subspecies in the M. abscessus complex consisting of M. abscessus, M. massiliense, and M. bolletii) is a rapid-growing nontuberculous mycobacterium that is commonly associated with human infections ranging from superficial skin and subcutaneous sepsis to chronic recurrent respiratory disease resembling pulmonary tuberculosis. This emerging pathogen shows multidrug resistance mediated through its innate mechanisms or through mutations that arise under the selective pressure of antibiotic use in clinical practice.¹

Linezolid is the first member of an entirely new class of synthetic antibacterial agents called the oxazolidinone antibiotics that inhibit an early reaction of bacterial protein synthesis by binding to the 23S ribosomal RNA (rRNA) of the 50S subunit. It was introduced in 2000 as an effective last-resort agent against Gram-positive bacteria that are resistant to other antibiotics, such as vancomycin-resistant staphylococci and enterococci. It is also recommended for the treatment of M. abscessus complex infections that are resistant to the more standard therapy by clarithromycin and amikacin.²

Disconcertingly, linezolid-resistant strains of M. abscessus have been documented in different parts of the world,^3,4 but the mechanisms driving these resistances in M. abscessus remain largely unknown. In this study, through the characterization and comparison of linezolid-resistant mutants, we aimed to identify the molecular determinants of linezolid resistance in M. abscessus.

Materials and Methods

Bacterial strains

The M. abscessus strains used in this study included a clinical strain M61, two selected mutants A2 and A2a(1), and two recombinant strains M61-fadD32mut and M61-pMV261 (Table 1). The procedures used to generate the mutant and recombinant strains are described in the following sections. In addition, 49 clinical isolates of M. abscessus complex were also included in this study for PCR analysis (Supplementary Table S1).

Table 1.

The Mycobacteroides abscessus Strains Used For The Characterization of Mutants

Strain	Plasmid	Linezolid MIC (mg/L)	Note
M61	N/A	0.25	A linezolid-susceptible clinical strain isolated from a sputum sample
A2	N/A	1	The first-step linezolid less-susceptible mutant derived from M61
A2a(1)	N/A	>256	The second-step linezolid-resistant mutantderived from A2
M61-fadD32mut	pMV261-fadD32mut	1	M61 transformed with the pMV261 plasmid carrying the mutant fadD32 gene (with c880t mutation)
M61-pMV261	pMV261	0.25	M61 transformed with the empty pMV261 plasmid

MIC, minimum inhibitory concentration; N/A, not applicable.

Mutant selection

The selection of spontaneous mutants was carried out using the method described previously.⁵ In brief, the wild-type M61 cells (linezolid minimum inhibitory concentration [MIC]: 0.25 mg/L) were inoculated into Middlebrook 7H9 broth (Becton Dickinson) for incubation at 37°C. On the 4th day of incubation, the broth culture (∼10⁸ CFU/mL) was plated onto nutrient agar (Becton Dickinson) containing 1 mg/L of linezolid (Cayman Chemical). A colony growing on the agar and confirmed to have a linezolid MIC of 1 mg/L was designated as the first-step mutant A2. This mutant was then cultured in broth as described for M61 and used for the selection of the second-step mutant, A2a(1), by plating onto agar containing 8 mg/L of linezolid.

Antimicrobial susceptibility testing

Stokes disk diffusion

Using cotton swabs, test and control strains, adjusted to the McFarland number 0.5, were plated onto the two halves of a cation-adjusted Mueller–Hinton agar plate (Isolab), leaving a gap of not more than 5 mm between the two strains. Using a pair of sterile forceps, an antibiotic-impregnated disk (Becton Dickinson) was placed at the center of the gap. The plate was incubated at 37°C for 3–5 days before the zones of inhibition were measured. Each test was performed in at least two biological replicates. Since the clinical breakpoints of antibiotic zone sizes have not been described for M. abscessus, we used the breakpoints of Staphylococcus aureus described in the leaflet of the Becton Dickinson BBL Sensi-Disc to interpret the results.

Etest

Etest (bioMérieux) was carried out to determine the MIC of linezolid, according to the manufacturer's protocol. MICs were interpreted according to the linezolid breakpoints described for rapid-growing mycobacteria (susceptible ≤8 mg/L, intermediate = 16 mg/L, resistant ≥32 mg/L) in the Clinical and Laboratory Standards Institute (CLSI) guideline.⁶

PCR analyses

PCR amplification was conducted on the 23S rDNA and fadD32 genes. For these assays, total DNA was purified from plate cultures using Quick-DNA Fungal/Bacterial DNA Miniprep (Zymo Research), according to the manufacturer's instructions. PCRs were prepared by mixing 1 × GoTaq Green Mastermix (Promega), 0.2 μM forward and reverse primers (Supplementary Table S2), nuclease-free water, and 10 ng of purified DNA.

PCR amplification was performed in the Veriti Thermal Cycler (Applied Biosystems) using the following conditions: 1 cycle of 95°C for 10 minutes, 40 cycles of 95°C for 30 seconds, annealing temperature (Supplementary Table S2) for 30 seconds, and 72°C for 60 seconds, and 1 cycle of 72°C for 10 minutes. A no-template control was included in every run. PCR products were resolved and visualized using gel electrophoresis. If necessary, PCR products were sent to Apical Scientific Ltd for Sanger sequencing.

Whole-genome sequencing

DNA samples of M61 and A2a(1) were submitted to Codon Genomics Ltd for PCR-free library preparations using the VAHTS Universal DNA Library Prep Kit for Illumina (Vazyme Biotech). The libraries were sequenced by Illumina HiSeq X Ten, using a 2 × 150 bp sequencing protocol.

The generated reads were preprocessed with Trimmomatic version 0.38.⁷ De novo assembly of the M61 reads was performed using SPAdes version 3.12.0.⁸ Clean paired-end reads of A2a(1) were mapped to the genome assembly of M61 using bowtie2 version 2.3.4.⁹ Variant-calling analysis was performed using the mpileup and bcftools algorithms of SAMtools version 0.1.19.¹⁰ The variant candidates were annotated using Annovar.¹¹ The sequencing data of M61 and A2a(1) were submitted to the European Nucleotide Archive (Accession No.: PRJEB38893).

Complementation

As the fadD32 gene in both A2 and A2a(1) contained the same mutation (c880t), only the fadD32 gene from A2 was selected for cloning. This gene was cloned into the Mycobacterium–Escherichia coli shuttle vector pMV261 (a kind gift from Dr. Therdsak Prammananan, National Center for Genetic Engineering and Biotechnology, Thailand)¹² using restriction enzymes (EcoRI and HindIII) and T4 ligase (New England Biolabs), as described previously.⁵ The recombinant plasmid was then electroporated into the M61 electrocompetent cells using Eporator (Eppendorf) to generate M61-fadD32mut, which was screened by colony PCR. M61-pMV261 was included as the empty-plasmid control.

Statistical analysis

The difference between the means of two groups was analyzed using the two-sample t-test in GraphPad Prism version 5.03.

Results and Discussion

In this study, we selected spontaneous mutants from the linezolid-susceptible M61 (MIC: 0.25 mg/L). The first-step mutant (A2) had reduced susceptibility to linezolid (MIC: 1 mg/L), but the second-step mutant, A2a(1), showed high level resistance to linezolid (MIC: >256 mg/L; Table 1). Comparative genomics analysis between A2a(1) and M61 revealed three mutations in the mutant genome, two of which were found in the 23S rDNA (g2244t and g2788t) and one in a gene encoding the fatty-acid-CoA ligase FadD32 (c880t → H294Y).

The PCR amplification of A2a(1) confirmed the presence of all three mutations, but that of A2 showed only the c880t mutation in fadD32. Complementation of the wild-type M61 with the pMV261 plasmid carrying the mutant fadD32 gene caused the previously sensitive M61 to develop the reduced susceptibility to linezolid (a change in MIC from 0.25 to 1 mg/L; Table 1). It is apparent from these results that the reduced susceptibility to linezolid in the first-step mutant A2 was due to the c880t mutation in the fadD32 gene, but the resistance in the second step mutant A2a(1) was largely due to the 23S rDNA mutations.

The 23S rRNA is the molecular target of linezolid in a bacterial cell and mutations in this gene may alter the binding sites of the drug, leading to the development of resistance. Using whole-genome sequencing, Ye et al. (2019) reported several 23S rDNA mutations in clinical isolates of M. abscessus complex that were resistant to linezolid.⁴ However, the mutations described by this group of authors did not include the g2244t and g2788t mutations found in this study. Interestingly, the g2244t and g2788t mutations of the 23S rDNA gene found in A2a(1) correspond exactly to the g2270t and g2814t mutations (Fig. 1), which have previously been associated with linezolid resistance in M. tuberculosis.¹³

FIG. 1.

Partial gene sequences of 23S rDNA. The sequences of Mycobacteroides abscessus M61 and A2a(1) were aligned to the sequence of M. tuberculosis H37Rv. The positions indicated in the red rectangles are according to the nucleotide positions of the 23S rDNA of H37Rv. The nucleotide positions 2270 and 2814 of M. tuberculosis correspond to nucleotide positions 2244 and 2788 of M. abscessus, respectively.

To assess the prevalence of these mutations among clinical isolates, we performed 23S rDNA PCR sequencing on 49 strains of M. abscessus complex (14 M. abscessus, 34 M. massiliense, and 1 M. bolletii) isolated from skin, subcutaneous and pulmonary specimens in a routine diagnostic laboratory. We did not observe the g2244t and g2788t mutations or the mutations described by Ye et al. (2019)⁴ in any of these isolates (Supplementary Table S1 and Supplementary Fig. S1). Phenotypically, these isolates were also not overtly resistant to linezolid, although 10 strains (6 M. abscessus, 3 M. massiliense, and 1 M. bolletii) showed intermediate resistance (MIC: 12 mg/L). Hence, the 23S rDNA mutations observed in A2a(1) appeared to be still rare among M. abscessus complex clinical isolates.

The FadD32 protein plays an important role in the formation of the outer membrane and in the cell-wall permeability of Mycobacterium.¹⁴ As fadD32 is an essential gene in mycobacteria,¹⁴ it is reasonable for us to believe that the fadD32 mutation is unlikely a loss-of-function mutation. Otherwise, the mutant would not be viable. Since FadD32 is involved in the biosynthesis of mycolic acids, which are known to limit the effectiveness of hydrophilic antibiotics,¹⁵ we performed Stokes disk diffusion to study the susceptibility patterns of A2 that harbored only the fadD32 mutation, against four lipophilic and four hydrophilic antibiotics from different classes (Table 2).

Table 2.

Antimicrobial Susceptibility Testing of A2 and M61 with Hydrophilic and Lipophilic Antibiotics

Antibiotic	Family	Hydrophilic/lipophilic	Average zone size (mm)		Significance
Antibiotic	Family	Hydrophilic/lipophilic	A2	M61	Significance
Imipenem	Carbapenem	H	33.5	38.5	^*
Linezolid	Oxazolidinone	H	30	40	^*
Vancomycin	Glycopeptide	H	15.5	45	^**
Amikacin	Aminoglycoside	H	32.5	33.5
Ciprofloxacin	Fluoroquinolone	L	25	24.5
Clarithromycin	Macrolide	L	41	42
Tigecycline	Glycylcycline	L	48.5	50.5
Clindamycin	Lincosamide	L	No zone	No zone

Two-sample t-test was used to analyze the difference; ^*p < 0.05, ^**p < 0.01.

The information about the chemical properties of the antibiotics (hydrophilic/lipophilic) was obtained from the following references.^18,19

Along with linezolid, A2 was found to be less susceptible to three of four hydrophilic antibiotics tested (imipenem, vancomycin, and linezolid). Unexpectedly, both A2 and M61 were equally susceptible to the hydrophilic amikacin (Table 2), suggesting that there may be alternative routes for this antibiotic to cross the cell wall of M. abscessus. In contrast, no difference in zone sizes was observed between A2 and M61 for three of three lipophilic antibiotics (clarithromycin, ciprofloxacin, and tigecycline). Clindamycin (lipophilic) was excluded from the comparison because both A2 and M61 did not produce any visible inhibition zones for this antibiotic. The apparent association of A2 with reduced susceptibility to hydrophilic antibiotics suggests that the c880t mutation in fadD32 may possibly alter the cell-wall permeability, limiting the entry of hydrophilic antibiotics.

Conclusion

FadD32 has been hailed as an attractive antimycobacterial drug target as it is genetically conserved among mycobacteria and it can be inhibited by many chemical analogs.¹⁶ In this study, our findings suggest an unexpected potential role of FadD32 in the development of antibiotic resistance. Although the c880t mutation in A2 was only associated with reduced susceptibility to linezolid and several other hydrophilic antibiotics, it is well documented that low-level antibacterial resistance in microbes is often the stepping stone to the development of high-level resistance of clinical significance.¹⁷ The FadD32 mutation might have contributed to the development of high-level resistance in the A2 offspring, A2a(1), which acquired two more mutations in the 23S rDNA gene that are also associated with linezolid resistance. To the best of our knowledge, the c880t (in fadD32), g2244t (in 23S rDNA), and g2788t (in 23S rDNA) mutations are new findings in M. abscessus.

In antibiotic resistance studies, mutants were normally generated from established type strains of a bacterial species. In this study, however, we were unable to select resistant mutants from ATCC 19977 (the type strain of M. abscessus) as it is already resistant to linezolid.³ Interestingly, ATCC 19977 has identical fadD32 and 23S rDNA sequences as our linezolid-sensitive M61, implying that the resistance phenotype of ATCC 19977³ is likely an outcome of other unknown genetic determinants.

In addition, cost considerations limited the whole-genome sequencing analysis to one linezolid-resistant mutant. Future endeavors should be made to characterize more resistant mutants of M. abscessus to have a better understanding of linezolid resistance mechanisms in this important pathogen. We also acknowledge that the antimicrobial susceptibility testing methods used in this study, Stokes disk diffusion and Etest, are not CLSI-endorsed testing methods for nontuberculous mycobacteria. Nevertheless, both of these methods demonstrated a consistent and progressive increase in resistance from M61 (zone size: 40 mm, MIC: 0.25 mg/L) to A2 (zone size: 30 mm, MIC: 1 mg/L) and A2a(1) (zone size: no zone, MIC: >256 mg/L), which was in line with the findings of their genotypic characterization.

Footnotes

Acknowledgment

The authors thank Dr. Therdsak Prammananan (National Center for Genetic Engineering and Biotechnology, Thailand) for providing them the pMV261 plasmid used in this study.

Authors' Contributions

Y.F.N. and H.F.N. conceived and designed the experiments. H.F.N. performed the experiments. H.F.N. and Y.F.N. wrote the article. Both authors read and approved the final version of the article.

Disclosure Statement

No competing financial interests exist.

Funding Information

Both authors were supported by the Malaysia Toray Science Foundation: 19/G583 (4417/N01) and UTAR: 6555/1N02.

Supplementary Material

References

Nessar

, Cambau

, Reyrat

J.M

, et al. 2012. Mycobacterium abscessus: a new antibiotic nightmare. J. Antimicrob. Chemother. 67:810–818.

Griffith

D.E.

, Aksamit

, a Brown-Elliott

, et al. 2007. An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases. Am. J. Respir. Crit. Care Med. 175:367–416.

Wallace

R.J.

, Brown-Elliott

B.A

, Ward

S.C

, et al. 2001. Activities of linezolid against rapidly growing mycobacteria. Antimicrob. Agents Chemother. 45:764–767.

, Xu

, Zou

, et al. 2019. Molecular analysis of linezolid-resistant clinical isolates of Mycobacterium abscessus. Antimicrob. Agents Chemother. 63:e01842-18.

Ng, H.F. 2019. Selection and Characterization of a Tigecycline-Resistant Mutant of Mycobacterium abscessus to Identify Possible Resistance Determinants. Selangor (states), Malaysia: Universiti Tunku Abdul Rahman.

Woods

G.L.

, Brown-Elliott

B.A

, Conville

P.S

, et al. 2011. Susceptibility Testing of Mycobacteria, Nocardiae, and Other Aerobic Actinomycetes; Approved Standard—2nd ed. Wayne, PA: Clinical and Laboratory Standards Institute.

Bolger

A.M.

, Lohse

, and Usadel

. 2014. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics, 30:2114–2120.

Bankevich

, Nurk

, Antipov

, A.A. et al. 2012. SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J. Comput. Biol. 19:455–477.

Langmead

, and Salzberg

S.L.

. 2012. Fast gapped-read alignment with Bowtie 2. Nat. Methods, 9:357–359.

10.

, Handsaker

, Wysoker

, et al. 2009. The sequence alignment/map format and SAMtools. Bioinformatics, 25:2078–2079.

11.

Wang

, Li

, and Hakonarson

. 2010. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res. 38:e164–e164.

12.

Stover

C.K.

, de la Cruz

V.F

, Fuerst

T.R

, J.E. et al. 1991. New use of BCG for recombinant vaccines. Nature, 351:456–460.

13.

, Liu

, Jiang

, et al. 2019. Characterization of linezolid-resistance-associated mutations in Mycobacterium tuberculosis through WGS. J. Antimicrob. Chemother. 74:1795–1798.

14.

Cortes

, Singh

A.K

, Reyrat

J.M

, et al. 2011. Conditional gene expression in Mycobacterium abscessus. PLoS One, 6:29306.

15.

Marrakchi

, Lanéelle

M.A

, and Daffé

. 2014. Mycolic acids: structures, biosynthesis, and beyond. Chem. Biol. 21:67–85.

16.

Galandrin

, Guillet

, Rane

R.S

, et al. 2013. Assay development for identifying inhibitors of the mycobacterial FadD32 activity. J. Biomol. Screen. 18:576–587.

17.

Baquero

2001. Low-level antibacterial resistance: a gateway to clinical resistance. Drug Resist. Updat. 4:93–105.

18.

Roberts

J.A.

, and Lipman

. 2009. Pharmacokinetic issues for antibiotics in the critically ill patient. Crit. Care Med. 37:840–851.

19.

Vitrat

, Hautefeuille

, Janssen

, et al. 2014. Optimizing antimicrobial therapy in critically ill patients. Infect. Drug Resist. 7:261–271.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

0.01 MB

0.17 MB