Unusually high clarithromycin resistance in Mycobacterium abscessus subsp. abscessus isolated from human gastric epithelium

Mycobacterium abscessus subsp. abscessus is a rapidly growing facultative intracellular pathogen that usually infects human lung and skin epithelium. Recently, we and another group have shown that it also has the potential to colonize human gastric epithelium, but its significance with respect to gastric diseases remains unclear. Although Helicobacter pylori still remains the only definite gastric pathogen, recent studies have shown that M. abscessus subsp. abscessus also has the potential to colonize human gastric epithelium. M. abscessus subsp. abscessus is known to exhibit multidrug resistance and clarithromycin has been used as the drug of choice. We aimed to determine the clarithromycin resistance profile of 117 (74 rough and 43 smooth) gastric M. abscessus subsp. abscessus strains and to detect the point mutations in rrl and erm (41) genes conferring the resistance. Our data showed 79.48% (19 smooth and 74 rough) of M. abscessus subsp. abscessus strains were resistant to clarithromycin (MIC90 ≤ 512 μg/mL), while 20.51% (24 smooth) were susceptible (MIC90 ≤ 8 μg/mL). Nucleotide sequence analysis of the rrl gene with reference strains of M. abscessus subsp. abscessus did not show any mutation that is relevant to the clarithromycin resistance. However, analysis of erm (41) gene showed that M. abscessus subsp. abscessus strains, which were susceptible to clarithromycin had C, C, G, and C at their nucleotide positions 28, 159, 238, and 330, respectively, while the resistant strains showed T, T, A, and A at the same positions. Based on antibiogram and sequence analysis data we recommend further studies involving genomic analysis to identify the other genes involved in high clarithromycin resistance in gastric M. abscessus subsp. abscessus along with the mechanisms involved.


Introduction
Mycobacterium abscessus subspecies abscessus (Mycobacterium abscessus subsp.abscessus) is a non-tuberculous mycobacteria (NTM) and is known for its rapid growth and resistance to multiple drugs.It is known to cause pulmonary infection and skin and soft tissues infections (mostly nosocomial) in humans (Lee et al., 2015).M. abscessus subsp.abscessus infection is difficult to treat because of its intrinsic resistance to most macrolide and other antibiotics including the classical anti-tuberculous drugs (Nessar et al., 2012;Griffith and Daley, 2022).
Recently, we and another group have isolated M. abscessus subsp.abscessus from human gastric epithelium.Interestingly, in Trivandrum, Kerala, India, the prevalence of gastric M. abscessus subsp.abscessus is even higher than the prevalence of Helicobacter pylori, a well-known gastric pathogen, which causes gastric cancer and peptic ulcer (Al-Momani et al., 2017;Chouhan et al., 2019).The common treatment regimen for H. pylori related gastric diseases is a proton pump inhibitor (e.g., Lansoprazole) and antibiotics.Because of the indiscriminate use of metronidazole to prevent amoebiasis in diarrhoea-endemic places in India and other countries, most H. pylori strains are resistant to metronidazole and clarithromycin is mostly the drug of choice against H. pylori (Safavi et al., 2016;Gonzales et al., 2019;Shetty et al., 2019).The significance of gastric colonization of M. abscessus subsp.abscessus with respect to gastric diseases is unknown at present, but the potential of this bacterium to cause diseases should not be neglected.Therefore, the resistance profile of the gastric M. abscessus subsp.abscessus strains against clarithromycin are worth studying for effective management of gastric diseases.
In 1990s, clarithromycin was the choice of drug to eradicate M. abscessus subsp.abscessus (Mushatt and Witzig, 1995;Brown-Elliott and Wallace, 2002).Clarithromycin was not effective against M. abscessus subsp.abscessus with point mutations at A 2058 G, C and A 2059 G, C (Escherichia coli numbering) or A 2270 → G or C and A 2271 → G or C (M. abscessus subsp.abscessus numbering) positions in the rrl gene that encodes the peptidyltransferese domain of 23S rRNA of bacterial ribosome (Wallace et al., 1996).Apart from mutations in rrl gene, another mechanism confers resistance against macrolides in NTM.The M. abscessus subsp.abscessus strains with functional erm (41) gene show an inducible resistance against clarithromycin upon prolonged incubation (14 days), while the M. abscessus subsp.abscessus strains with non-functional erm (41) gene show susceptibility to clarithromycin.In the erm (41) gene, a substitution of T-to-C at position 28 (T 28 C), which leads to the alteration of Trp to Arg at the 10 th amino acid of the peptide, was found to be associated with loss of function and susceptibility to clarithromycin (Nash et al., 2009).The isolates of M. abscessus subsp.massiliense have not shown any inducible macrolide resistance because of a 397-bp deletion in erm (41) gene, which results in a non-functional erm (41) gene (Kim et al., 2010).
The aim of this study was to determine the minimum inhibitory concentration (MIC) of the gastric M. abscessus subsp.abscessus strains against clarithromycin and to understand the genetic basis of the resistance.

Ethics statement
The study was approved by the Institute Human Ethics Committee of Rajiv Gandhi Centre for Biotechnology (Approval Number IHEC/01/2017/18) and by the Human Ethics Committee of Govt. Medical College, Trivandrum (Approval Number IEC.No.05/07/2016/ MCT).Patients between the age of 20 and 70 years were recruited for the study and written informed consents were obtained from all patients.Trivandrum is in the southern part of India and is the capital city of Kerala, mostly the part of western ghats with high humidity.

Mycobacterium abscessus subsp. abscessus culture and characterization
A total of 117 M. abscessus subsp.abscessus (rough and smooth) strains isolated from human gastric biopsies were used in this study.The M. abscessus subsp.abscessus strains were grown on Brain Heart Infusion (BHI) agar plates containing calf serum (7%) and were incubated at 37°C in microaerobic conditions (5% O 2 , 10% CO 2 , 85% N 2 ).M. abscessus subsp.abscessus colonies were identified at the species level based on the growth rate, colony morphology (rough and smooth), and pigmentation as well as 16S rRNA gene sequence analysis.Partial nucleotide sequencing of the hsp65 gene was used for further confirmation of gastric M. abscessus subsp.abscessus strains to distinguish them from closely related M. bolletti, M. chelonae and M. massiliense.Phylogenetic analysis was done by using BioEdit software (version 7.2.6.1).

Antibiotic susceptibility
M. abscessus subsp.abscessus strains were tested for clarithromycin (macrolide antibiotic) susceptibility (from 0.125 μg/mL to 512 μg/mL) by agar dilution and broth microdilution assay.BHI plates were prepared using newborn calf serum (7%) and required concentrations of antibiotics, and BHI broth was prepared similarly for the microdilution assay and finally supplemented with antibiotic as per the desired concentration.After 3, 7, and 14 days of incubation with clarithromycin in the microwell, 10 μL liquid culture from each treated well was applied on the BHI plates and were incubated in microaerobic incubator for 3-7 days.To verify the inducible resistance of gastric M. abscessus subsp.abscessus against clarithromycin, gastric M. abscessus subsp.abscessus were pre-treated with clarithromycin (0.1 μg/mL) for 3 days and then the MIC was determined.M. abscessus subsp.abscessus clarithromycin breakpoint (MIC 90 > 8 μg/mL) was determined according to the Clinical and Laboratory Standards Institute (CLSI) guideline published in 2011.Broth microdilution-based methodology for antimicrobial susceptibility testing of nontuberculous mycobacteria has been considered the gold standard (Woods et al., 2011).

Bacterial DNA isolation
The bacterial DNA was isolated as previously described (Berg et al., 1997).In brief, the bacterial colonies were harvested in 500 μL PBS and 10.3389/fmicb.2023.1193380Frontiers in Microbiology 03 frontiersin.orgcentrifuged at 5,000 rcf for 10 min.The bacterial pellet was resuspended in 200 μL GTE (glucose/tris/EDTA) buffer.The bacterial suspension was then treated with lysozyme (10 mg/mL) at 37°C for 1 h.After enzymatic digestion, bacterial cells were lysed using TES (tris/EDTA/ SDS) buffer.Proteinase K (50 μg/mL) and RNase (20 μg/mL) were then added and the tubes were incubated at 55°C for 2 h.The digested bacterial proteins were removed by phenol: chloroform: isoamyl alcohol and then by chloroform: isoamyl alcohol treatments.The bacterial DNA was precipitated using 3 M sodium acetate (pH 5.2) and chilled absolute ethanol.The precipitated DNA was washed with 70% ethanol and the dried pellet was dissolved in 1X TE (tris-EDTA) buffer of pH 8.

Colony morphologies and clarithromycin resistance patterns of the gastric Mycobacterium abscessus strains
The M. abscessus subsp.abscessus strains isolated from individuals with various gastric diseases have two distinct colony morphologies: smooth and rough.A total of 117 gastric M. abscessus subsp.abscessus (74 rough and 43 smooth) strains were tested for clarithromycin resistance using agar dilution and broth microdilution based assays.The MIC 90 for gastric M. abscessus subsp.abscessus rough morphotypes was ≤256 μg/mL after 14 days of incubation, while M. abscessus subsp.abscessus rough morphotypes grown in 0.1 μg/mL clarithromycin exhibited an induced increase in MIC 90 showed MIC 90 of ≤512 μg/mL after 14 days of treatment (Table 1).Similarly, clarithromycin treatment (uninduced and induced) was carried out for the 24 smooth M. abscessus subsp.abscessus morphotypes.We observed that all 24 M. abscessus subsp.abscessus smooth morphotypes had MIC 90 ≤ 4 μg/mL after 14 days of incubation in uninduced conditions, while in induced conditions, the M. abscessus subsp.abscessus smooth morphotypes showed MIC 90 ≤ 8 μg/ mL after 14 days of incubation (Table 1)

Mycobacterium abscessus subsp. abscessus sensitive and resistant strains
In order to investigate the molecular basis of clarithromycin resistance in gastric M. abscessus subsp.abscessus strains, a 670 bp region of the erm (41) gene was amplified by PCR.Once erm (41) gene was amplified from M. abscessus subsp.abscessus (rough and smooth) resistant and sensitive strains, the amplified products were visualized on 1.5% of agarose gel.An additional ~180 bp amplicon was observed only for the sensitive strains along with the expected band of 670 bp and the results were consistent for all total 117 strains irrespective of smooth and rough morphotypes (Figure 1A).The additional low molecular weight amplicon (180 bp) has not been reported previously.However, with a different set of primer targeting the erm (41) gene, only a single amplicon of 764 bp was observed (Figure 1B).

Sequence analysis of the erm (41) and rrl genes
As mentioned above, we obtained two amplicons (670 bp and 180 bp) in sensitive and one amplicon (670 bp) in resistant strains.The amplicons were purified and sequenced separately to confirm the association of clarithromycin resistance profiles with singlenucleotide polymorphism (SNP) in the erm (41) gene.Earlier reports suggest that SNP at 28 (C to T) nucleotide position is associated with inducible resistance of M. abscessus subsp.abscessus.We observed all clarithromycin susceptible M. abscessus subsp.abscessus strains (Pfister et al., 2004) had nucleotide C (GenBank accession number MW147114) and all resistant strains (93) strains had nucleotide T (GenBank accession number MW147113 and MW147115) at the position 28 of erm (41) gene (Table 2; Figure 2A).Along with the T 28 C SNP, we also observed nucleotide C at position 159 (T 159 C), nucleotide G at position 238 (A 238 G), and nucleotide C at position 330 (A 330 C) in all susceptible M. abscessus strains (GenBank accession number MW147114), but these mutations were absent in all resistant strains of M. abscessus (GenBank accession number MW147113 and MW147115), irrespective of smooth and rough morphotypes.The erm (41) genes nucleotide sequences were converted to amino acid sequences for both resistant and susceptible M. abscessus subsp.abscessus strains and was compared with reference strains.We observed Arginine (Arg) and Valine (Val) at position 10 and position 80, respectively only in susceptible strains, while for all resistant strains, Tryptophan (Trp) and Isoleucine (Ile) were observed which is similar to the reference strain (Table 2; Figure 2B).We also amplified rrl gene of gastric M. abscessus subsp.abscessus and sequenced the nucleotides to confirm SNPs in rrl gene but we did not observe any SNPs at 2270 A to G or C and 2271 A to G or C (M. abscessus numbering) in the rrl gene of M. abscessus resistant (GenBank accession number MW148480 and MW148478) and susceptible strains (GenBank accession number MW148479) as mentioned in Table 2; Supplementary Figure S1.

Sequence analysis of the low molecular weight amplicon of Mycobacterium abscessus subsp. abscessus
The low molecular weight amplified product (180 bp) of oxidoreductase gene from M. abscessus subsp.abscessus susceptible strains were sequenced using forward and reverse primers.As shown in Supplementary Figures S3A,B, the chromatograms of Sanger sequencing for both forward and reverse primers showed very distinct peaks.
Multiple sequence alignment did not show any match of these sequences with the erm (41) gene of M. abscessus reference strain as well as the gastric M. abscessus subsp.abscessus strains.To find out the identity of these sequences sequence homology analysis was performed using the BLAST algorithm on the NCBI platform.The BLAST analysis confirmed the identity of these sequences is not erm (41) gene but oxidoreductase gene (GenBank: CP029073.1 and CU458896.1) of M. abscessus strain G122 and M. abscessus ATCC19977.Oxidoreductase of M. abscessus contains a total of 3,552 nucleotides, which encode a 1,183 amino acid containing protein.As it was shown in Figure 3, the 180 bp amplicon shows 100 percent similarity with M. abscessus G122 strain, starting from G_2991 to C_3158 (GenBank accession number MW142321) which covers 167 bp (Figure 3A).The amino acid sequence starts from valine at the 133 position and the match ends at amino acid valine at the 187 position, which covers a total of 54 amino acids (Figure 3B).With respect to the reference strain M. abscessus ATCC19977, the oxidoreductase gene of the gastric M. abscessus subsp.abscessus showed 4 point mutations at C 3007 → G, C 3042 → T, C 3109 → G and A 3135 → G (Supplementary Figure S2).Altogether, these results confirmed that the low molecular weight (180 bp) band obtained in the erm (41) PCR is amplified from the oxidoreductase gene present in gastric M. abscessus subsp.abscessus strains.

Discussion
M. abscessus subsp.abscessus is a rapidly growing Mycobacterium species responsible for pulmonary and soft tissue infections.Although rare, the bacteria also cause disseminated infection, especially in immunocompromised individuals (Lee et al., 2015).M. abscessus subsp.abscessus is a non-tuberculous mycobacteria (NTM) showing a high level of antibiotic resistance and poses a serious challenge to disease management (Nessar et al., 2012).Various antibiotics which are found to be effective against M. abscessus subsp.abscessus, like azithromycin, amikacin, meropenem, ciprofloxacin, imipenem, trimethoprim/sulfamethoxazole, and clarithromycin.Among these antibiotics, clarithromycin remains to be the drug of choice to treat M. abscessus subsp.abscessus related infections (Griffith et al., 2007;  Nessar et al., 2012).Recently we reported the presence of M. abscessus subsp.abscessus in gastric epithelium of patients with various gastric diseases (Chouhan et al., 2019).These gastric diseases are commonly associated with H. pylori infection.Recommended treatment option for the eradication of H. pylori and the management of related gastric diseases is standard clarithromycin-based triple therapy.In this study, we determined the clarithromycin resistance patterns of gastric M. abscessus subsp.abscessus strains.
The clarithromycin susceptibility for gastric M. abscessus subsp.abscessus strains were determined by clarithromycin broth microdilution assay and agar dilution method, where the bacteria was incubated in various concentrations of clarithromycin for 3, 7, and  Frontiers in Microbiology 07 frontiersin.org14 days followed by inspecting bacterial viability.After incubation for 14 days, the growth (viability) pattern of the susceptible and resistant strains were determined.The microdilution-based antibiotic susceptibility test has been recommended by CLSI, because of its reproducibility (Reller et al., 2000;Woods et al., 2011).The clarithromycin susceptibility data obtained in our study of 117 gastric M. abscessus subsp.abscessus strains demonstrate that the phenotype of clarithromycin susceptibility was fully concordant with erm (41) gene SNPs.Our data suggest that of the 117 M. abscessus subsp.abscessus strains, 93 (79.48%) strains were resistant to clarithromycin, while 24 (20.51%)strains were susceptible to clarithromycin.
The point mutation (C to T) at position 28 of erm (41) gene is known to be associated with inducible clarithromycin resistance by M. abscessus subsp.abscessus.M. abscessus subsp.abscessus strains showed susceptibility at day 3 of treatment but gradually acquired resistance after 7 days to 14 days of incubation (Nash et al., 2009;Kim et al., 2010;Bastian et al., 2011).Surprisingly we observed there were 19 smooth M. abscessus subsp.abscessus strains which were resistant to clarithromycin and had wild-type erm (41) genotype as like M. abscessus subsp.abscessus rough morphotype, while susceptible M. abscessus subsp.abscessus smooth morphotype had SNPs in their erm (41) gene.We observed after 14 days of incubation M. abscessus  Mutation from isoleucine (Ile) to valine (Val) at the 80 th position has been associated with macrolide drug resistance in M. abscessus strains isolated from Korea.Nash et al. has shown that Erm41 protein with Trp10 was associated with resistance while Arg10 was associated with susceptible phenotype, as the protein harbouring Arg10 was non-functional (Nash et al., 2009;Lee et al., 2014).Our amino acid sequence analysis for Erm (41) protein suggests that all susceptible strains of M. abscessus subsp.abscessus had Arginine (Arg) amino acid at 10 th position, while Tryptophan (Trp) was present in resistant strains of M. abscessus subsp.abscessus.We also observed mutation at position 80, where Valine (Val) was replaced by Isoleucine (Ile) in resistant strains of gastric M. abscessus subsp.abscessus.Our erm (41) gene analysis for gastric M. abscessus subsp.abscessus strains correlated with susceptibility pattern of M. abscessus subsp.abscessus strains irrespective of smooth and rough morphotypes.The sequence of low molecular weight (180 bp) amplicon did not show any homology with erm (41) gene of M. abscessus strains but BLAST analysis identified that 180 bp amplicon belongs to oxidoreductase gene of M. abscessus strain G122 with 100 percent match.On the other hand, when compared with M. abscessus ATCC19977 strain, the 180 bp amplicon sequence displayed 4 point mutations, C 3007 → G, C 3042 → T, C 3109 → G and A 3135 → G in oxidoreductase gene (Supplementary Figure S2).Most of the antibiotics also exert their bactericidal effect by generating reactive oxygen species (ROS) or targeting bacterial redox systems.The bacterial oxidoreductase gene has been linked to antibiotic resistance through neutralizing toxic molecules, detoxification of antibiotics, and repair damage caused by antibiotics.Recent studies have revealed that M. abscessus subsp.abscessus induces efflux pump encoding genes in response to antibiotic stress specifically antibiotic targeting ribosome (Egorov et al., 2018;Mudde et al., 2022;Schildkraut et al., 2022).A mutation in the oxidoreductase gene leading to a non-functional oxidoreductase enzyme may exert a detrimental effect on bacteria during antibiotic treatment.In this study, we observed a possible link between the oxidoreductase gene of M. abscessus subsp.abscessus in clarithromycin resistance.Further study involving whole genome sequencing of gastric M. abscessus subsp.abscessus strains are needed to understand the antibiotic resistance gene pool present in the strains along with genotypes and the nature of resistance (acquired or induced).To conclude, we determined the clarithromycin resistance profile of the gastric M. abscessus subsp.abscessus strains and studied the genetic basis of clarithromycin resistance.We have also demonstrated induced clarithromycin resistance in the isolated gastric M. abscessus subsp.abscessus strains.Our finding also suggests that the C to T mutation at the 28 th nucleotide position of the erm (41) gene has an important role in conferring clarithromycin resistance and erm (41) gene sequence analysis can light on the mechanism of clarithromycin resistance in of gastric M. abscessus subsp.abscessus.The results of our study will be helpful while designing efficient strategies to combat multi-drug resistant strains of M. abscussus subsp.abscessus as well as for the management of associated gastric diseases.
. Based on clarithromycin sensitivity pattern we have 3 different types of gastric M. abscessus subsp.abscessus strains-(a) resistant M. abscessus subsp.abscessus rough (Mabs-R) (b) resistant M. abscessus subsp.abscessus smooth type A (Mabs-S-A) and (c) susceptible M. abscessus subsp.abscessus smooth type B (Mabs-S-B) by considering MIC 90 ≤ 8 μg/mL as a cut-off in induced as well as uninduced conditions as per the Clinical and Laboratory Standards Institute (CLSI) guideline published in 2011.

FIGURE 3
FIGURE 3Sequence alignment of M. abscessus subsp.abscessus strains with low molecular weight amplicon.(A) Gene sequence of low molecular weight amplicon shows 100 percent similarity with the oxidoreductase gene of M. abscesus subsp.abscessus strain G122 chromosome.(B) low molecular weight amino acid sequence showed 100 percent similarity with oxidoreductase protein.
(Pfister et al., 2004;Lipworth et al., 2018)f 256 μg/mL when they were not induced.However, upon induced with clarithromycin, the MIC raised to 512 μg/mL of clarithromycin.MabsS-A strains were susceptible to 256 μg/mL of clarithromycin in uninduced conditions and MIC was 512 μg/mL in induced conditions.On the other hand, MabsS-B showed MIC of 8 μg/mL after induction, when compared to 4 μg/mL of clarithromycin in uninduced conditions.We observed resistant phenotype with wild type erm (41) gene with T, T, A, and A nucleotide base at the 28, 159, 238, and 330 positions, respectively but strains with C, C, G, and C at the 28, 159, 238, and 330 nucleotide positions respectively, exhibited susceptible phenotype (Tables1, 2).Acquired and inducible resistance for clarithromycin in M. abscessus has already been reported.Acquired resistance for clarithromycin has been associated with point mutations at A 2270 → G or C and A 2271 → G or C (M. abscessus numbering system) of rrl gene(Pfister et al., 2004;Lipworth et al., 2018).To our surprise, we did not observe any mutations in gastric M. abscessus subsp.abscessus strains at position A 2270 → G or C and A 2271 → G or C (M. abscessus numbering system) of rrl gene.In conclusion, no association was observed between rrl gene and clarithromycin MIC for the resistant and sensitive phenotype of gastric M. abscessus subsp.abscessus strains. subsp