Characterization of an MDR Lactobacillus salivarius isolate harbouring the phenicol-oxazolidinone-tetracycline resistance gene poxtA Free

Abstract

Objectives

To characterize the oxazolidinone resistance gene poxtA in a Lactobacillus salivarius isolate of pig origin.

Methods

L. salivarius isolate BNS11 was investigated for the presence of mobile oxazolidinone resistance genes by PCR. Antimicrobial susceptibility testing was performed by broth microdilution. Transfer experiments were conducted to assess horizontal transferability of the gene poxtA. WGS was carried out using a combination of Oxford Nanopore MinION/Illumina HiSeq platforms. The presence of translocatable units (TUs) carrying resistance genes was studied by PCR assays and subsequent sequence analysis.

Results

L. salivarius isolate BNS11 was positive for poxtA. WGS showed that it harboured two gene copies each of the poxtA and the fexB genes, which were located on the broad-host-range Inc18 plasmid pBNS11-37kb and in the chromosomal DNA, respectively. The plasmid-borne poxtA gene together with the genes fexB, vat(E) and erm(C) were located in an MDR region on plasmid pBNS11-37kb. Analysis of the genetic context showed that an approx. 11 kb poxtA-fexB fragment was integrated into the chromosomal DNA and two novel IS elements ISLasa1 and ISLasa2 were identified in this inserted fragment. PCR assays revealed that five different IS1216E-based TUs carrying the resistance genes poxtA, fexB, vat(E) or erm(C) were formed.

Conclusions

To the best of our knowledge, this is the first report of the transferable oxazolidinone resistance gene poxtA in the genus Lactobacillus. In addition, the presence of IS1216E-based TUs will contribute to the persistence and accelerate the dissemination of resistance genes, including poxtA.

Introduction

Oxazolidinones are highly effective antimicrobial agents for treating severe infections caused by multiresistant Gram-positive bacteria, including MRSA and VRE.¹ However, the emergence and spread of mobile oxazolidinone resistance genes, including cfr, optrA and poxtA,² pose a serious therapeutic challenge to human medicine. Among them, the gene poxtA codes for a ribosomal protection protein of the ATP-binding cassette (ABC)-F family, which confers decreased susceptibility or resistance to oxazolidinones, phenicols and tetracyclines.³ Since its first description in an MRSA strain from Italy, the gene poxtA has been described in 11 countries on four continents.⁴ However, this gene has been identified exclusively in Enterococcus spp. and Staphylococcus spp. so far.⁴ The genus Lactobacillus is a member of the lactic acid bacteria group and part of the commensal gut microbiota of humans and animals.⁵ Among them, the species Lactobacillus salivarius is usually regarded as a candidate probiotic.⁶ However, the frequent detection of antimicrobial resistance genes (ARGs) in Lactobacillus poses a challenge to its use as a probiotic.⁵ In this study, we identify a poxtA-carrying porcine L. salivarius isolate and investigate the genetic context and potential transmission mechanisms of poxtA.

Materials and methods

Bacterial strain and species identification

During a routine surveillance study, a total of 300 faecal samples were randomly collected from healthy pigs in a pig farm in Heilongjiang province, China in 2021. All samples were incubated in buffered peptone water for 12–16 h at 37°C and streaked on brain heart infusion (BHI) agar containing 4 mg/L linezolid. Through screening and Gram staining, 34 (34/300, 11.3%) Gram-positive strains that grew on these plates were identified. Species identification was conducted using 16S rRNA sequencing as previously described.⁷ Among the 34 strains, 33 belonged to the genus Enterococcus, while the remaining strain (1/34, 2.9%) proved to be an L. salivarius, designated BNS11. This strain was further investigated in this study.

Antimicrobial susceptibility testing (AST)

The MICs were determined by broth microdilution as previously described.⁸ EFSA guidelines⁹ were used to interpret the results. For rifampicin and lincomycin, previously proposed interpretive criteria were adopted.⁸ Both Staphylococcus aureus ATCC 29213 and Streptococcus pneumoniae ATCC 49619 served as the quality control strains.

Investigation of mobile oxazolidinone resistance genes

The presence of mobile oxazolidinone resistance genes was detected by PCR assays as previously described.¹⁰ The PCR products were subjected to Sanger sequencing for further confirmation.

Transfer experiments

Conjugation and electrotransformation assays were carried out using the rifampicin- and fusidic-acid-resistant recipient strain Enterococcus faecalis JH2-2 as previously described.⁵ BHI plates were supplemented with 50 mg/L rifampicin, 25 mg/L fusidic acid and 8 mg/L florfenicol and served for screening of the transconjugants and transformants. Transformants were confirmed by PCR assays and AST.

WGS and bioinformatics analysis

WGS was performed using a combination of Illumina HiSeq and Nanopore MinION platforms. The short- and long-read data were subjected to de novo hybrid assembly using Unicycler tool v0.4.3.¹¹ Genome annotation was performed using the RAST web server (http://rast.nmpdr.org) combined with BLAST analysis. ARGs and plasmid replicon types were identified using the CGE server (https://cge.cbs.dtu.dk/services/). The circular map of the plasmid was generated using BRIG tool v.0.95.

Detection of translocatable units (TUs)

The presence of IS1216E-mediated TUs was detected by PCR assays using the primer pairs listed in Table S1 (available as Supplementary data at JAC Online). The resulting amplicons were then subjected to Sanger sequencing to obtain the complete sequences of the TUs.

Nucleotide sequence accession numbers

The complete genome of strain BNS11 was deposited in the GenBank database under accession numbers CP089850–CP089856.

Results and discussion

Characterization of the strain BNS11

The L. salivarius strain BNS11 exhibited an MDR phenotype and relatively high MICs of linezolid and florfenicol (Table S2). PCR results showed that this strain was poxtA positive. WGS revealed that strain BNS11 contained chromosomal DNA of 1 865 220 bp and six plasmids, including pBNS11-360kb (359 787 bp), pBNS11-37kb (37237 bp), pBNS11-29kb (28 852 bp), pBNS11-3kb-1 (2958 bp), pBNS11-3kb-2 (2883 bp) and pBNS11-2kb (2037 bp). ResFinder analysis identified two copies of the poxtA gene, which were located on plasmid pBNS11-37kb and in the chromosomal DNA, respectively.

Characteristics of the poxtA-carrying plasmid pBNS11-37kb

Plasmid pBNS11-37kb consisted of 30 predicted ORFs with an average G + C content of 33.99% (Figure 1a). BLAST searches did not identify significant similarity matches when using the complete sequence of plasmid pBNS11-37kb as the query. PlasmidFinder analysis showed that plasmid pBNS11-37kb contained the rep_{2_pRE25} replication gene (repR) and was assigned to the broad-host-range Inc18 plasmid family. Based on a previous study,⁴ it is assumed that plasmids of the Inc18 family are involved in the dissemination of poxtA across different Gram-positive genera and species. The characteristics of the Inc18-family plasmid pBNS11-37kb in the genus Lactobacillus further support this assumption. In addition to the poxtA gene, plasmid pBNS11-37kb also carried the resistance genes erm(C), fexB and vat(E) (Figure 1a).

Conjugation assays showed that plasmid pBNS11-37kb is non-conjugative. However, transformants carrying plasmid pBNS11-37kb were obtained by electrotransformation. AST results indicated that the transformants displayed elevated MICs of linezolid, chloramphenicol, florfenicol, erythromycin and lincomycin in comparison with the recipient strain (Table S2). Interestingly, the introduction of poxtA into the recipient failed to increase the tetracycline MIC (Table S2). A similar observation has been made in a previous study¹² and is likely due to the fact that poxtA is not a true tetracycline resistance gene, but only slightly increases the tetracycline MICs below the threshold for resistance.⁴

Analysis of the genetic context revealed that the plasmid-borne poxtA gene was part of an MDR region and was bracketed by two directly oriented IS1216E copies (Figure 1b). As previously reported, the poxtA gene is most frequently found together with four small ORFs encoding hypothetical proteins flanked by IS1216E copies in the same orientation.⁴ However, the poxtA-bearing segment between the two IS1216E copies on pBNS11-37kb lacked three of the four ORFs for hypothetical proteins (Figure 1b).

Genetic environment of the chromosomal poxtA gene

As shown in Figure 2, the poxtA gene was adjacent to a fexB-bearing area bracketed by two directly oriented IS1216E copies. In the poxtA-fexB segment, two novel IS-like elements were found. They were submitted to the ISfinder database and received the designations ISLasa1 and ISLasa2. Furthermore, two direct target site duplications of 3 bp, 5′-TAT-3′ and 5′-CTC-3′, were found immediately upstream and downstream of both elements, respectively, which might represent the signature of a transposition event. Sequence analysis showed that ISLasa1 was inserted into the 3′-end of the poxtA gene, thereby disrupting the poxtA gene into two parts. BLAST searches revealed that, except for the two novel IS elements, the poxtA-fexB segment was highly similar to the corresponding region of plasmid pC27-2 (GenBank accession number MH784602) and plasmid pC25-1 (GenBank accession number MH784601), both from Enterococcus faecium. The flanking regions of the ∼11 kb poxtA-fexB segment, which included the genes hp and rec upstream of the ΔpoxtA gene to the ΔyibE/F gene downstream of the second IS1216E, exhibited striking homology with the corresponding region of the chromosomal DNA of L. salivarius strain ZLp4b (GenBank accession number CP062071). This observation implied that the poxtA-fexB segment was integrated into this chromosomal region of strain BNS11, thereby replacing the genes between rec and yibE/F and causing a truncation of the gene yibE/F (Figure 2). Compared with plasmid-borne poxtA genes, the occurrence of chromosomal poxtA genes is, so far, a rare observation.⁴ Considering the similarity of the poxtA-fexB segment to that of plasmid pC27-2, it is possible that the chromosomal poxtA-fexB segment originated from an enterococcal plasmid and that the IS elements ISLasa1 and ISLasa2 have subsequently inserted into this segment.

Detection of TUs carrying ARGs

As shown in Figure 1(a), the plasmid pBNS11-37kb harboured six copies of IS1216E, all of which were located in the same orientation. In addition, two IS1216E elements flanking the chromosomal fexB gene were also located in the same orientation (Figure 2). Recently, it has been shown that, similar to IS26, the IS1216E element can mobilize ARGs via the formation of TUs.¹³ To investigate whether IS1216E-mediated TUs were generated, PCR assays and Sanger sequencing were performed. Sequence analysis showed that a total of five TUs carrying ARGs, TU1–5, were formed from both the plasmid pBNS11-37kb and the chromosomal DNA (Figures 1b and 2). Since the sequences of the two IS1216E copies 4 and 5 were identical, it is impossible to determine whether IS1216E copy 4 or copy 5 was present in TU4. This observation demonstrated that IS1216E is a highly active genetic element. It is noteworthy that these IS1216E-derived TUs can then integrate either into plasmids or in the chromosomal DNA by recombination with another IS1216E copy.¹⁴ In this way, the ARGs, that are part of the respective TUs, were able to move between different chromosomal and plasmidic locations.

Conclusions

To the best of our knowledge, this study provides the first description of poxtA in Lactobacillus. Of note, the co-occurrence of the poxtA gene with other ARGs on the broad-host-range Inc18-family plasmid pBNS11-37kb may lead to the co-selection and dissemination of the poxtA gene. The formation of TUs will contribute to the persistence and dissemination of ARGs. Attention should be paid to the potential risks of the transfer of the poxtA gene from L. salivarius to other Gram-positive pathogens.

Funding

This work was supported by the Natural Science Foundation of Heilongjiang Province of China (YQ2019C031) and the German Federal Ministry of Education and Research (BMBF) under project number 01KI2009D as part of the Research Network Zoonotic Infectious Diseases.

Transparency declarations

None to declare.

Supplementary data

Tables S1 and S2 are available as Supplementary data at JAC Online.

References