Molecular Characterization of OXA-198 Carbapenemase-Producing Pseudomonas aeruginosa Clinical Isolates

Clinical histories of the four cases are summarized in Table 1. Hospital A (first case in 2010) and hospital B (three later cases in 2012 and 2013) are located 120 km from each other, with no histories of patient transfer between the two institutions. Remarkably, none of the patients had traveled abroad. Isolates 5301-1 and 5301-2 were recovered in the same intensive care unit (ICU) and during the same time period, suggesting the occurrence of cross-transmission between the two patients. In contrast, the third patient (isolate 5301-4) was transferred from another Belgian hospital more than 6 weeks after the death of patient 5301-2, thus excluding the occurrence of direct patient-to-patient cross-transmission. The origin of colonization for the third patient remains unknown; it could have been due to a contaminated respiratory device or other possible environmental reservoirs.

All four P. aeruginosa isolates were resistant to penicillins and penicillin-inhibitor combinations and displayed decreased susceptibility to carbapenems (Table 2). They were also resistant to aminoglycosides and fluoroquinolones (highlighting their phenotype of multidrug resistance), but they remained susceptible to expanded-spectrum cephalosporins (ceftazidime [except 5301-4] and cefepime [PA41437]), to the ceftazidime-avibactam combination, and to colistin. The isolates 5301-4 and PA41437, which were already more resistant to expanded-spectrum cephalosporins, were also resistant to ceftolozane-tazobactam. All four strains belonged to serotype O11. Multilocus sequence typing (MLST) analysis of the four isolates revealed an allelic profile for acs, aro, gua, mut, nuo, pps, and trp of alleles 18, 4, 5, 3, 1, 17, and 13, respectively, which corresponded to sequence type 446 (Table 3), a type already reported for patients from different continents and clinical origins, as deposited in the PUBMLST database.

Together, these findings indicated that all four isolates were clonally related. Single-nucleotide polymorphism (SNP) analysis revealed that isolates 5301-1 and 5301-2 were identical, whereas 1 SNP was identified in 5301-4 and 31 SNPs were identified in PA41437, in comparison to 5301-1. These 31 SNPs could be explained by the fact that PA41437 was isolated in 2010 and the three other isolates were isolated in 2013. Interhospital spread of a single OXA-198-producing P. aeruginosa strain could not be evidenced by epidemiological data, since there was a time interval of 3 years, a distance of 120 km between the two hospitals, and an absence of any patient exchange or transfer between the two hospitals. It should be highlighted that, besides the three isolates reported here, no other OXA-198-producing P. aeruginosa isolates were identified in a multicenter survey conducted in 67 Belgian hospitals in 2013 (196 isolates intermediate/resistant to carbapenems and resistant to at least three classes of antibiotics) (Y. Glupczynski, unpublished data). Furthermore, in a similar and more recent Belgium multicenter survey performed in 2016 (163 multidrug-resistant, carbapenem-intermediate/resistant P. aeruginosa isolates obtained from patients from 56 different hospitals), no single OXA-198-producing P. aeruginosa isolate could be evidenced.

Besides the presence of the bla_OXA-198 gene, the resistome of the four tested isolates revealed the presence of several resistance genes, including a catB7-like gene (98.4% nucleotide identity with catB7), qacEΔ1, and sul1 (conferring resistance to quaternary ammonium and sulfonamides, respectively) (Table 3). These resistance traits were carried by the plasmid pOXA-198 (Fig. 1A). Immediately downstream of the bla_OXA-198 gene, a copy of aacA7 coding for AAC(6′)-I was detected. Other resistance determinants were identified from the whole-genome annotation, including a novel variant of the naturally occurring oxacillinase gene bla_OXA-50-like (3 amino acid changes) and the chromosomal cephalosporinase gene bla_PAO, as well as a fosA-like gene (99.0% identity with fosA), cmlA1, and a aph(3′)-IIb gene. Substitutions in topoisomerases were also investigated, to identify potential mutations that could explain the fluoroquinolone resistance. Typical substitutions in the quinolone-resistance-determining region (QRDR), i.e., T83I in GyrA and S87L and A587T in ParC, were identified. To explain the high MICs for imipenem, mutations in the OprD2-porin-coding genes causing alterations or modifications were investigated. A deletion of 88 nucleotides, encompassing the 5′ end of the oprD2 gene and the ribosomal binding site (RBS), was identified; the deletion resulted in loss of the RBS and the first 9 amino acids of OprD2, suggesting the absence of the OprD2 porin.

Sequencing yielded 20,490 high-quality reads, giving rise to 4,395,424 bp, which were derived from the pOXA-198 library. De novo assembly, performed using CLC Workbench 10.0.1 (Qiagen, Courtaboeuf, France), yielded 31 contigs, of which only 17 were derived from pOXA-198. After gap closures and verification of the position of each contig by PCR, the overall structure of pOXA-198 included a replication, a partition, partial conjugation, and accessory modules (Fig. 1A). This plasmid was untypeable using PCR-based replicon typing (PBRT) targeting enterobacterial plasmids and did not correspond to any of the 10 known IncP-type plasmids commonly identified in Pseudomonas species (11); therefore, we propose to name it IncP-11. Research on synteny was performed using Mauve software (Fig. 1B). This plasmid shared common features with pRSB101 (GenBank accession number AJ698325) and pRSB105 (GenBank accession number DQ839391), including the replication module. These two plasmids were identified from uncultured bacteria in activated sludge (12, 13). BLAST searches of GenBank revealed that IncP-11 was found on plasmid pRSB105 along with IncP-6 and on plasmid pRSB101, which harbored only IncP-11. This IncP-11 replicon confers the ability to replicate in P. aeruginosa, in Escherichia coli, and in Xanthomonas campestris (12). The partitioning system associated with the replicase showed high levels of identity with partitioning systems discovered in environmental bacteria such as Xanthomonas, Aeromonas, and Gulbenkiania species, thus suggesting a likely environmental origin for these IncP-11-bearing plasmids. Several genes showing homology with transfer operons have been identified, but a complete transfer operon is absent from this plasmid. This could explain why this plasmid could not be transferred by conjugation, as described previously (10).

Concerning the accessory genes identified within the plasmid sequence, the bla_OXA-198 gene is embedded within a class 1 integron, together with the catB7-like gene (10). Several remnant structures of transposons, i.e., Tn402-like and TnAS3-like, both belonging to the Tn3 family of transposons, have also been identified within the pOXA198 sequence. These two transposons are likely to be inactive due to partial deletions. As for partitioning and replication systems, these not yet fully characterized transposons shared homology with transposons identified in soil bacteria. Another mobile element, similar to ISPa38, which was identified previously in P. aeruginosa, was also identified. The element present on plasmid pOXA-198 shares with ISPa38 only the presence of tnpA, the likely transposase gene, and it lacks the seven other open reading frames present in ISPa38 (14). The remaining part of the plasmid contains mainly hypothetical proteins with unknown function, which need to be characterized in more detail.

Mapping the reads obtained from whole-genome sequencing (WGS) of the four OXA-198-producing P. aeruginosa isolates against pOXA-198 revealed that the plasmids from the four isolates displayed 100% nucleotide sequence identity.

We described here the molecular features of the bla_OXA-198-harboring plasmid. This plasmid belonged to a novel IncP type that we named IncP-11, based on sequence homology. The transfer operon was partly deleted, explaining why the transfer by conjugation failed. This could limit the spread of this plasmid and favor clonal diffusion.

Genome analysis of the four OXA-198-producing P. aeruginosa isolates revealed that the isolates were clonally related. However, neither a reservoir nor any common source of transmission between the patients harboring P. aeruginosa OXA-198 isolates 5301-1 and 5301-2 and the last isolate (isolate 5301-4) could be detected. These isolates could be considered to belong to a cluster episode (rather than an outbreak) and could possibly reflect interregional spread of this clone, although this could not be demonstrated. Remarkably, only 31 SNPs were identified between the first isolate and the three later isolates, which were collected from patients in two different and distantly located hospitals without any obvious link or context. Overall, this study suggests that the spread of the bla_OXA-198 gene is linked to a single P. aeruginosa clone rather than dispersion of a plasmid, which is not self-transferable.

P. aeruginosa isolate PA41437 was originally isolated in 2010 from several lower respiratory tract specimens from a patient with acute exacerbation of chronic bronchitis who had been admitted to hospital A, as described previously (10). Three additional P. aeruginosa clinical isolates (5301-1, 5301-2, and 5301-4) were recovered from clinical specimens from patients hospitalized at hospital B in Belgium in 2013. The three OXA-198-positive P. aeruginosa strains were collected in the setting of a multicenter survey on antimicrobial resistance in P. aeruginosa that was organized by the Belgian national reference center in early 2014. The P. aeruginosa reference strain PAO1, P. aeruginosa PU21, and Escherichia coli TOP10 (Invitrogen, Merelbeke, Belgium), harboring the natural plasmid pOXA-198, were used (10, 15, 16).

Bacterial species identification was confirmed by matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) with a Microflex LT mass spectrometer (Bruker Daltonik GmbH, Leipzig, Germany). In vitro determination of MICs for 12 antimicrobial agents was performed with the microdilution method, using customized Sensititre plates (BEGN3B; Trek Diagnostic Systems), and results were interpreted according to CLSI breakpoints (17).

Specific primers were used to detect the most common carbapenemase genes, as described previously (10). Detection of the bla_OXA-198 gene was performed using primers OXA-41437FW (5′-CTCGAATTCATGCATAAACACATGAGTAAG-3′) and OXA-41437RV (5′-CTCAAGCTTTTATTCGATCCCCCTTT-3′) (10).

O serotyping was performed for all four isolates with the slide agglutination test, using polyvalent antisera and 16 monovalent antisera (designated O1 to O16), according to the manufacturer's instructions (Bio-Rad, Marnes-La-Coquette, France).

Plasmid DNA from P. aeruginosa PA41437 was extracted using a plasmid maxiprep kit (Qiagen), according to the manufacturer's instructions. Total DNA was extracted from colonies using an UltraClean microbial DNA isolation kit (Mo Bio Laboratories, Ozyme, Saint-Quentin, France), according to the manufacturer's instructions. The DNA concentrations and purity were controlled with a Qubit 2.0 fluorometer, using double-stranded DNA high-sensitivity and/or broad-range assay kits (Life Technologies) (18).

The DNA library for the plasmid extracted from E. coli strain TOP10 (pOXA-198) (2 by 250-bp paired-end reads) and those for the whole genomes of the four P. aeruginosa isolates (2 by 150-bp paired-end reads) were prepared using Nextera XT kits v2 (18). The libraries were run on a HiSeq automated system (Illumina, Paris, France). Reads were trimmed to remove poor-quality sequences (Q values of ≥20, with a word size of 34). De novo assembly was performed using CLC Workbench 9.0.1 (Qiagen). The acquired antimicrobial resistance genes were identified using the Resfinder server v2.1 (http://cge.cbs.dtu.dk/services/ResFinder-2.1) (19). MLST was performed in silico using the MLST server (https://cge.cbs.dtu.dk/services/MLST/), and SNP tree construction was performed using CSI phylogeny (https://cge.cbs.dtu.dk/services/CSIPhylogeny/) (20, 21). The plasmid and genomes were annotated using the RAST server (22).

Contigs from plasmid pOXA-198 were considered DNA contamination if they showed at least 90% similarity with E. coli K-12 substrain MG1655 genomic DNA sequences (GenBank accession number NC_000913.2), using the BLASTn algorithm. Gap closures and verification of the position of each contig were performed by PCRs. Circular representation was obtained using DNA plotter (http://www.sanger.ac.uk/science/tools/dnaplotter) (23). Alignment to observe the synteny of pOXA-198 was performed using Mauve software (http://darlinglab.org/mauve/mauve.html) (24).

Hypothetical natural plasmids were extracted and analyzed by electrophoresis on a 0.7% agarose gel, as described previously (18, 25). They were introduced into E. coli TOP10 by electroporation, using a Gene Pulser II (Bio-Rad). Recombinant clones were selected on tryptic soy agar supplemented with 50 μg/ml ticarcillin (Sigma, St. Quentin Fallavier, France).

The nucleotide sequence data reported in this paper have been deposited in the EMBL/GenBank nucleotide sequence database under accession numbers MG958650 for pOXA-198 and NMPW00000000, NMPV00000000, NMPU00000000, and NMPT00000000 for P. aeruginosa PA41437, 5301-1, 5301-2, and 5301-4, respectively.