Emergence of Tn1999.7, a New Transposon in bla_OXA-48-Harboring Plasmids Associated with Increased Plasmid Stability

ABSTRACT

OXA-48 is the most common carbapenemase in Enterobacterales in Germany and many other European countries. Depending on the genomic location of bla_OXA-48, OXA-48-producing isolates vary in phenotype and intra- and interspecies transferability of bla_OXA-48. In most bacterial isolates, bla_OXA-48 is located on one of seven variants of Tn1999 (Tn1999.1 to Tn1999.6 and invTn1999.2). Here, a novel Tn1999 variant, Tn1999.7, is described, which was identified in 11 clinical isolates from 2016 to 2020. Tn1999.7 differs from Tn1999.1 by the insertion of the 8,349-bp Tn3 family transposon Tn7442 between the lysR gene and bla_OXA-48 open reading frame. Tn7442 carries genes coding for a restriction endonuclease and a DNA methyltransferase as cargo, forming a type III restriction modification system. Tn1999.7 was carried on an ~71-kb IncL plasmid in 9/11 isolates. In one isolate, Tn1999.7 was situated on an ~76-kb plasmid, harboring an additional insertion sequence in the plasmid backbone. In one isolate, the plasmid size is only ~63 kb due to a deletion adjacent to Tn7442 that extends into the plasmid backbone. Mean conjugation rates of the Tn1999.7-harboring plasmids in J53 ranged from 4.47 × 10⁻⁵ to 2.03 × 10⁻², similar to conjugation rates of other pOXA-48-type IncL plasmids. The stability of plasmids with Tn1999.7 was significantly higher than that of a Tn1999.2-harboring plasmid in vitro. This increase in stability could be related to the insertion of a restriction-modification system, which can promote postsegregational killing. The increased plasmid stability associated with Tn1999.7 could contribute to the further spread of OXA-48.

INTRODUCTION

The prevalence of carbapenemase-producing Enterobacterales (CPE) is increasing worldwide due to dissemination of CPE clones and the spread of carbapenemase genes via mobile genetic elements. OXA-48 carbapenemases are the most frequently detected carbapenemases in Western Europe, including Germany, North Africa, and the Middle East (1). The spread of OXA-48 is mainly linked to the dissemination of a highly conjugative 63.6-kb IncL plasmid (2), referred to as pOXA-48. Within this plasmid, bla_OXA-48 is located between two copies of IS1999 in variants of the composite transposon Tn1999 (Fig. 1). As the transfer inhibition gene tir is disrupted by the insertion of Tn1999, pOXA-48 plasmids exhibit increased conjugation rates, which contributes to the spread of OXA-48 CPE (3). The most frequently identified Tn1999 variant is Tn1999.2, followed by Tn1999.1 and invTn1999.2, whereas the other variants are rarely detected (4). The transposon type can influence the phenotype of the host strain as well as mobilization characteristics of bla_OXA-48. For example, some transposon variants alter the antibiotic resistance phenotype of the isolate by the insertion of another β-lactamase, as described for Tn1999.4 (5), or increased expression of bla_OXA-48 by modification of the promoter region, as described for Tn1999.2 (6). On the other hand, invTn1999.2 has been associated with chromosomal integration of bla_OXA-48 (7).

FIG 1.

Overview of Tn1999 variants located on pOXA-48-type plasmids (5–7, 42–45). Arrows indicate coding sequences and their directions: green arrows, Δtir; brown arrows, Δaac (acetyl-CoA carboxylase); blue arrows, (Δ)lysR; red arrows, bla_OXA-48; gray arrows, genes related to mobile genetic elements; orange arrows, genes related to a restriction modification system; light red arrows, other genes. Gray and black boxes, insertion sequences and transposons; yellow triangles, target site duplications; IRL/IRR, inverted repeat left/right. The figure style was adapted from Pitout et al. (1).

This study analyzed the genetic background of a novel Tn1999 variant, termed Tn1999.7, identified in clinical isolates of different OXA-48-producing Enterobacterales species. Furthermore, we aimed to investigate possible influences of Tn1999.7 on conjugation and plasmid stability, as well as the resistance phenotype of the host strain.

RESULTS AND DISCUSSION

Epidemiology.

Between 2013 and 2020, 139 OXA-48-like CPE were collected at the University Hospital Cologne. After removing isolates with non-OXA-48 carbapenemases (e.g., OXA-181) and with bla_OXA-48 not located on IncL plasmids, 72 strains remained for further characterization. In these isolates, PCR identified 11 isolates that appeared to have an unknown Tn1999 variant. Genomic analysis by whole-genome sequencing (WGS) confirmed a novel Tn1999 variant, termed Tn1999.7, in 10 isolates, and a truncated version, termed ΔTn1999.7, in one isolate (Table 1). Tn1999.7 and ΔTn1999.7 have not been described before, and no similar sequences could be retrieved from GenBank at the time of writing.

TABLE 1.

Overview of clinical isolates harboring Tn1999.7 or ΔTn1999.7

Isolate	Isolation mo/yr	Patient ID	Species	ST	OXA-48 plasmid size (bp)	Transposon	Additional plasmid(s) (size in bp)	Genome accession no.
KA14695	10/2016	01	Klebsiella aerogenes	168	71,167	Tn1999.7	Untypeable (137,841)	JAKLSS000000000
CF14695	10/2016	01	Citrobacter freundii	22	71,166	Tn1999.7	None	JAKLST000000000
EC3239	03/2017	01	Escherichia coli	10	75,519	Tn1999.7	Untypeable (98,178), IncFIC (139,536), IncX1 (51,546)	JAKLSU000000000
EC13927	09/2018	02	Escherichia coli	58	71,166	Tn1999.7	IncFII (143,810)	JAKLSV000000000
RO13927	09/2018	02	Raoultella ornithinolytica	105	71,166	Tn1999.7	None	JAKLSW000000000
CF3626	09/2018	03	Citrobacter freundii	116	71,166	Tn1999.7	IncC (104,041), untypeable (90,108), IncFII (54,153)	JAKLSX000000000
CF17067	10/2018	04	Citrobacter freundii	22	63,126	ΔTn1999.7	Untypeable (136,836)	JAKLSY000000000
KP17051	12/2018	04	Klebsiella pneumoniae	15	71,166	Tn1999.7	IncFIB (109,959), IncR (37,365)	JAKLSZ000000000
CF15807	11/2018	05	Citrobacter freundii	19	71,166	Tn1999.7	IncFII (130,575), IncFIB (62,797)	JAKNAE000000000
EC8448	06/2019	06	Escherichia coli	453	71,166	Tn1999.7	IncFIB (183,507), p0111 (136,595), IncFII (52,179)	JAKLTA000000000
CF3910	02/2020	07	Citrobacter freundii	19	71,166	Tn1999.7	IncFII (130,677), IncFIB (60,937)	JAKLTB000000000

Conventional multilocus sequence typing revealed the same sequence type (ST22 or ST19) in two Citrobacter freundii isolates each. Core genome single nucleotide polymorphism (SNP) analysis of CF14695 and CF17067 (both ST 22) showed four SNPs and a 3-bp insertion difference between the isolates, which were obtained within 24 months of each other. For CF15807 and CF3910, sharing ST19, the difference was three SNPs and they were isolated 15 months apart. For other Enterobacterales species, a mutation rate of approximately 3 × 10⁻⁷ nucleotides per site per year has been reported (8). Assuming similar rates for C. freundii, which has a genome size of ~5 million bp, a genome-wide SNP rate of about 1.5 per year would be expected, not considering mutation-enhancing environments or genetic recombination. Thus, it cannot be excluded that any of the pairs sharing the same ST had a common source. However, with at least 7 of 11 Tn1999.7-carrying bacterial isolates being different species, the spread of Tn1999.7 appears to be plasmid driven rather than clonal.

Genomic analysis of transposon variants.

Genomic analysis of the WGS hybrid assemblies showed that Tn1999.7 is equivalent to Tn1999.1, with the addition of an 8,349 Tn3 family transposon (designated Tn7442 by the Transposon Registry [9]) inserted between bla_OXA-48 and lysR (Fig. 1). Tn7442 appears to be a hybrid transposon, in which the sequence between the resolution site res1 of tnpR and the right inverted repeat is equal to Tn1, including the Tn1 transposase (tnpA). Downstream of the resolution site, Tn7442 comprises a gene coding for a putative resolvase, showing 87% amino acid identity with the Tn2 resolvase (tnpR) (accession no. ALT06231). The resolution site res1 is a typical recombination site for Tn3 family transposons (10). In contrast to Tn1, which carries bla_TEM-2 as cargo, Tn7442 appears to carry genes encoding a type III restriction modification (R-M) system. This is composed of two open reading frames (ORFs) showing 99% sequence identity in a REBASE BLAST search with the putative restriction subunit Kgr39ORF30055P (accession no. QLP11302) and the putative DNA methyltransferase subunit M.SmaBWH35ORF26500P (accession no. AVU43264), respectively, with both enzymes sharing the target sequence CCGCAG. From GenBank, plasmid sequences containing Tn7442 could be retrieved for Klebsiella pneumoniae, Hafnia paralvei, and Citrobacter portucalensis but were from different Inc types and were not associated with bla_OXA-48 (11, 12). The presence of the Tn7442 insertion in Tn1999.7 was confirmed by PCR in all isolates. Isolate CF17067 showed a variant of Tn1999.7, termed ΔTn1999.7, in which the sequence between the left inverted repeat of Tn7442 and repA of the plasmid backbone was deleted (Fig. 1 and 2).

FIG 2.

Comparison of the plasmid structures of pOXA-48.7, pOXA-48.7c, and pOXA-48.7b. Red, bla_OXA-48; blue, Tn7442; gray, insertion sequences; green, additional insertion sequence in the pOXA-48.7b plasmid backbone; orange, genes of the pOXA-48 plasmid backbone or hypothetical proteins.

Analysis of Tn1999.7-carrying plasmids.

Tn1999.7 was located on a 71,166-bp plasmid of the IncL group, termed pOXA-48.7, in 9/11 isolates (Fig. 2). Outside the Tn1999.7 transposon, the plasmid is highly similar to pOXA-48a (13), but with the insertion of IS1R downstream of the korC gene as described before for other pOXA-48 variants (7). The plasmid of EC3239 had an additional 4,353-bp insertion upstream of IS1R. This insertion encompasses another copy of IS1R and three ORFs that match sequences from C. freundii-associated plasmids in GenBank. This plasmid, termed pOXA-48.7b, was 75,519 bp (Table 1 and Fig. 2). The deletion in the plasmid of CF17067 (see the paragraph above) led to a smaller plasmid size of 63,126 bp (Fig. 2). This plasmid was termed pOXA-48.7c.

Horizontal gene transfer in vitro.

To assess the influence of Tn1999.7 on the transferability of the pOXA-48-type plasmids in vitro, liquid mating assays were performed with sodium azide-resistant Escherichia coli J53 and Klebsiella quasipneumoniae PRZ as recipients. Conjugation of the bla_OXA-48-carrying IncL plasmid into J53 was successful for all Tn1999.7 isolates and into PRZ for all Tn1999.7 isolates except for Klebsiella aerogenes KA14695. However, the plasmid from the latter strain could be transferred to PRZ via transformation. For C. freundii CF17067, conjugation of the plasmid carrying the truncated variant ΔTn1999.7 to either recipient strain was unsuccessful. Nevertheless, the plasmid could be transferred by transformation into J53 but not into PRZ. The inability of horizontal gene transfer via conjugation observed in ΔTn1999.7 in vitro is most likely caused by the missing trb region. Genes of the trb operon are essential for assembling the mating pair apparatus. Therefore, the spread of pOXA-48.7c is likely compromised, which is supported by the fact that ΔTn1999.7 was found in only one patient.

Conjugation rates of the bla_OXA-48-carrying IncL plasmids in J53 were quantified for all Tn1999.7 isolates and compared to isolates carrying plasmids that share the same plasmid backbone, but carrying bla_OXA-48 on Tn1999.1, Tn1999.2, a Tn1999.2 variant, and invTn1999.2. For the Tn1999.7-carrying isolates, mean conjugation rates ranged from 4.47 × 10⁻⁵ to 2.03 × 10⁻², with a mean of means of 3.77 × 10⁻³ (see Fig. S1 in the supplemental material). Mean conjugation rates of the control group ranged from 1.13 × 10⁻⁴ to 2.81 × 10⁻², with a mean of means of 6.22 × 10⁻³. These results show no significant difference in the conjugation rates between the two groups. Even though species-specific comparisons between the two groups have only limited significance due to small species groups, the results also indicate no apparent difference in plasmid conjugation rates between the same species that host Tn1999.7 and other Tn1999 variants. However, it appears that a higher plasmid number of the donor isolate correlates with a lower conjugation rate in this study. Inhibitory effects of coresiding plasmids are more common than facilitating effects (14). However, a deeper analysis of these correlations is beyond the scope of this study.

Evaluation of plasmid contents of four representative J53 transconjugants (Tc_J53EC3239, Tc_J53CF3626, Tc_J53KP17051, and Tc_J53EC2013) via long-read sequencing showed that all transconjugants solely contained the pOXA-48-type plasmid. This plasmid could be further conjugated from the J53 transconjugants into another recipient, E. coli DOTN, indicating that no additional plasmids are needed for mobilization.

The antibiotic susceptibility phenotype was determined for the clinical isolates, transconjugants, and transformants (Table S1). Carbapenem MICs of transconjugants/transformants carrying Tn1999.7 and Tn1999.2 showed no significant differences.

Plasmid stability.

In Tn1999.7, an R-M system was inserted via the transposon Tn7442. As effects of R-M systems on plasmid stability have been described (15), plasmid stability was assessed using two different J53 transconjugants carrying pOXA-48.7 (three runs each) or pOXA-48.7b. The stability of these plasmids was compared with that of a J53 transconjugant carrying pOXA-48.2, the most frequently detected Tn1999.2-carrying plasmid. At 37°C, pOXA-48.7 was significantly more stable than pOXA48.2 at day 7 (95.7% [SD, 8.1%] versus 79.9% [SD, 11.8%]) and day 14 (90.9% [SD, 24.2%] versus 27.5% [SD, 24.2%]) (Fig. 3). At 42°C, plasmid stability was lower for pOXA-48.2 and pOXA-48.7 at day 14. Still, pOXA-48.7 was significantly more stable at this incubation temperature than pOXA-48.2 (day 7, 99.0% [SD 11.1%] versus 55.6% [SD 41.7%]; day 14, 67.4% [SD 33.0%] versus 15.4% [SD 17.5%]) (Fig. 3). The plasmid stability of pOXA-48.7b was on a similar level to that of pOXA-48.7 at both temperatures.

FIG 3.

Plasmid stability of the Tn1999.7-carrying plasmids pOXA-48.7 and pOXA-48.7b and the Tn1999.2-carrying plasmid pOXA-48.2 in J53. Each symbol represents the mean result from one liquid culture plated in duplicate at the respective time point. Asterisks indicate significant differences between pOXA-48.7 and pOXA-48.2.

Restriction-modification systems have been linked to increased plasmid stability (15). This is due to their ability to promote postsegregational killing of bacterial cells missing the plasmid with the R-M system, similar to a toxin-antitoxin system (16). In our study, the potential function of the type III R-M system correlates well with the observation of an increased plasmid stability in the R-M-carrying plasmids pOXA-48.7 and pOXA-48.7b. Nevertheless, definitive conclusions of the function must be drawn with caution as R-M system-related increase of plasmid stability has not been described for type III R-M system yet, but only for type II R-M systems (16). Other potential functions of type III R-M systems (i.e., expression-level changes due to a modified methylome [17]) have not been studied here, which is a limitation that should be addressed in future studies. However, when MICs of antibiotics were compared, no major changes could be observed between J53 transconjugants carrying Tn1999.7 and those carrying Tn1999.2, indicating a similar bla_OXA-48 expression.

In summary, we describe the novel bla_OXA-48-carrying transposon variant Tn1999.7, which is associated with increased plasmid stability in vitro, presumably due to an inserted R-M system. The increased plasmid stability associated with Tn1999.7 could contribute to the further spread of OXA-48, which will be revealed by future surveillance data.

MATERIALS AND METHODS

Bacterial isolates and antimicrobial susceptibility testing.

Enterobacterales isolates carrying bla_OXA-48-like genes were obtained between 2013 and 2020 from clinical samples obtained at the University Hospital Cologne. They were detected during routine diagnostics due to elevated carbapenem MICs, and bla_OXA-48 was verified via the automated PCR system Xpert Carba-R (Cepheid, Sunnyvale, CA, USA) (18). Only one isolate per species for each patient was included in the study. Antibiotic susceptibility was determined by the Micronaut-S broth microdilution panel (Merlin, Bornheim, Germany). Due to a small MIC range for carbapenems in this panel, susceptibility testing was complemented by gradient tests for ertapenem, meropenem, and imipenem (Liofilchem, Roseto degli Abruzzi, Italy).

PCR-based characterization of bla_OXA-48-harboring isolates.

DNA was extracted using the DNeasy UltraClean microbial kit (Qiagen, Hilden, Germany). Tn1999.1, Tn1999.2, and invTn1999.2 were identified by PCR as described before (19). For isolates carrying Tn1999.7, the insertion of Tn7442 was verified by PCR with a long-range high-fidelity polymerase (Phusion; ThermoFisher, MA, USA) with primers Tn_OXA-48-F and TN_TIR (19), resulting in amplicons of 3,512 bp for Tn1999.1, 3,508 bp for Tn1999.2, and 11,862 bp for Tn1999.7.

Whole-genome sequencing and bioinformatic analyses.

Whole-genome sequencing (WGS) of each sample with Illumina and Oxford Nanopore technology was performed at the Quantitative Biology Center of the University of Tübingen as previously described (20). Sequence data were processed with an in-house pipeline using Snakemake 5.31.1 (21), a python-based script language. Raw reads were quality trimmed with Trimmomatic 0.39 (22), Porechop 0.2.4 (23), and NanoFilt 2.7.1 (24). The quality of the trimmed reads was checked with FastQC 0.11.9 (25), MultiQC 1.9 (26), and NanoPlot 1.32.1 (24). Hybrid assemblies were created with Unicycler 0.4.8 (27), which uses Bowtie2 and Pilon for polishing. Centrifuge 1.0.4_beta (27) was used for species assignment. Annotation and characterization were done by comparing assembled sequences to the databases ResFinder (28), PlasmidFinder (29), and PubMLST (30) and an in-house transposon database with the software ABRicate 1.0.1 (31) and mlst 2.19.0 (32). Core genome single nucleotide polymorphism (SNP) analysis was calculated from raw reads using snippy 4.6.0-1 (33). Annotation from reference plasmids and further analysis was done manually with the software Geneious Prime (Biomatters, Ltd., Auckland, New Zealand). Open reading frames were analyzed via the BLAST function of NCBI (blastx), UniProt (34), ISfinder (35), and the REBASE database (36).

Analysis of horizontal gene transfer.

Qualitative conjugation was assessed using the sodium azide-resistant strains E. coli J53 and K. quasipneumoniae PRZ as recipient strains (37, 38). Quantitative conjugation was assessed using J53 as the recipient according to a previously published protocol (19) for all Tn1999.7-carrying isolates and compared to seven isolates with other Tn1999 variants. Briefly, liquid mating of clinical isolates and the recipient strains was performed for 2 h, followed by plating on coliform chromogenic agar (CCA) (Carl Roth, Karlsruhe, Germany) containing 20 μg/mL amoxicillin-clavulanate (Hexal, Holzkirchen, Germany) alone or in combination with 100 μg/mL sodium azide. Transconjugation frequency was determined after 48 h of incubation at 37°C by dividing the number of transconjugants (i.e., the number of colonies growing on agar with amoxicillin-clavulanate and sodium azide) by the number of donors (i.e., the number of transconjugants minus the colonies growing on amoxicillin-clavulanate agar without sodium azide).

To assess if J53 transconjugants were able to transfer the plasmid to other recipients, further qualitative conjugation experiments were carried out using representative J53 transconjugants and nalidixic acid-resistant E. coli DOTN as the recipient. CCA supplemented with 128 mg/L nalidixic acid and 20 mg/L amoxicillin-clavulanate was used for selection of DOTN transconjugants.

In case in vitro conjugation was unsuccessful, a transformation assay was employed. Plasmid DNA was extracted from clinical isolates utilizing the Plasmid Maxi kit (Qiagen). Transformation of electrocompetent J53 and PRZ bacteria was carried out as described before (39).

Plasmid stability.

The stability of plasmids carrying Tn1999.7 was measured using E. coli J53 transconjugants. They were compared to a J53 transconjugant from the above-mentioned isolate collection, with a Tn1999.2-carrying plasmid that has the same plasmid backbone as the Tn1999.7-carrying plasmids, further referred to as pOXA-48.2. Bacteria were cultured in an antibiotic-free medium for 14 days, and the loss of the resistant population was monitored. A suspension of transconjugants corresponding to a McFarland standard of 1.0 was diluted 1:10⁵, and 10 μL was subsequently added to 4 mL of 1/3 lysogeny broth (LB) (40). Liquid cultures were incubated under agitation (180 rpm) at 37°C, which best reflects the environment in the human gut, and 42°C, to simulate a more challenging environment for the bacteria. Every 24 h, bacterial suspensions were diluted 1:2 × 10⁵ in normal saline, and 10 μL of the dilution was used to inoculate 4 mL of fresh LB. On days 7 and 14, serial dilutions of the liquid cultures were plated on LB agar (LBA) and LBA containing 20 mg/L amoxicillin-clavulanic acid (LBA-AMC), each in duplicate. After overnight culture, the mean number of CFU was determined per duplicate. The ratio of colonies growing on LBA-AMC compared to antibiotic-free LBA was determined to calculate the percentage of bacteria carrying the pOXA-48 plasmid.

Statistics and graphs.

Plasmid conjugation and stability rates were compared with the Mann-Whitney U test. A P value of <0.05 was considered significant. Graphical representations for quantitative conjugation and plasmid stability rates were created with GraphPad Prism 8.4.3. Plasmid graphs were created with Easyfig 2.2.5 (41).

Ethical approval.

Isolates were collected from routine diagnostics, and only pure cultures of bacterial isolates were investigated. No patient-related data were analyzed. No ethical approval is necessary for this type of study according to the regulations of the University of Cologne.

Data availability.

The complete nucleotide sequence assemblies of all isolates harboring Tn1999.7 were deposited publicly in NCBI under BioProject no. PRJNA801874. Accession numbers are listed in Table 1.