The increased incidence of multidrug-resistant (MDR) Gram-negative bacteria worldwide over the last decade has been worrisome (Bassetti et al., 2019). Carbapenems are considered the drug of choice in treating such infections. However, the increased usage of these antibiotics has led to the emergence of carbapenem-resistant strains (Roberts et al., 2020). Center for Disease Control (CDC) considers carbapenem-resistant Enterobacterales (CRE) as a serious global threat to patient health that limits treatment options, especially in chronically ill patients in intensive care units (ICU) and long-term care facilities (McConville et al., 2017; Gupta et al., 2019). Resistance to carbapenems is caused by various mechanisms, such as porin loss, increased efflux pump activity and most importantly the production of carbapenemases (Ye et al., 2018).

Acquired carbapenem-hydrolyzing β-lactamases are enzymes of the Ambler class A KPC type, class B type, including IMP-, VIM-, and NDM-like metallo-β-lactamases (MβLs), or the class D OXA-48 type. However, in Enterobacterales, the most clinically significant metallo-β-lactamases are the NDM-like enzymes (Nordmann et al., 2011). NDM-like enzymes efficiently hydrolyze broad range of β-lactam antibiotics, including penicillins, cephalosporins, and carbapenems, with the exception of monobactams such as aztreonam (Nordmann and Poirel, 2014). In 2008, NDM-1 was reported for the first time (Yong et al., 2009) and, shortly after, 24 distinct NDM enzymes have been described worldwide, most of them originating from India, China, Nepal, or Near East (Hornsey et al., 2011; Kaase et al., 2011; Nordmann et al., 2012; Williamson et al., 2012; Rogers et al., 2013; Tada et al., 2013; Wang et al., 2014; Wu et al., 2019). Compared to NDM-1, NDM-5 MβL has two amino acid substitutions (Val88Leu and Met154Leu) (Hornsey et al., 2011). NDM-5 was reported, for the first time, from a clinical Escherichia coli strain in the United Kingdom (Hornsey et al., 2011). In the Czech Republic, NDM-1 and NDM-5 enzymes were reported for the first time in 2011 and in 2016, respectively (Hrabak et al., 2012; Paskova et al., 2018).

Also, OXA-48-producing Enterobacterales pose an important public threat, mainly due to their challenging detection and the rapid horizontal transfer of pOXA-48-like plasmids (Skalova et al., 2017). OXA-48-like enzymes hydrolyze penicillins at a high level and carbapenems at a low level, sparing broad-spectrum cephalosporins, and are not susceptible to β-lactamase inhibitors (Poirel et al., 2004). Since its first report in Turkey in 2004, 11 variants with few amino acid substitutions or deletions emerged globally (Bakthavatchalam et al., 2016; Mairi et al., 2018). In 2006–2007, OXA-181 was reported for the first time in India and since then it is considered one of the most disseminated OXA-48-like enzymes worldwide especially in patients with travel history to the Indian continent (Castanheira et al., 2011; Rojas et al., 2017). In 2011, OXA-244 was reported for the first time in Spain, since then there has been only limited number of reports, indicating limited dissemination (Oteo et al., 2013; Fursova et al., 2015; Potron et al., 2016; van Hattem et al., 2016). Recently, co-production of NDM- and OXA-48-like carbapenemase has been increasingly described, especially in patients with travel history to Italy (Marchetti et al., 2019), South Korea (Baek et al., 2019), Turkey (Otlu et al., 2018), Singapore (Balm et al., 2013) and the United States (Doi et al., 2014; Contreras et al., 2020).

Thus, the aim of this study was to genomically characterize four isolates (three Klebsiella pneumoniae and one E. coli) co-producing NDM- and OXA-48-like carbapenemases from Czech patients with travel history or/and previous hospitalization abroad.

All isolates expressed resistance to ampicillin, ciprofloxacin, piperacillin, piperacillin-tazobactam, cefotaxime, meropenem, and ertapenem, while retaining susceptibility to colistin. Moreover, all isolates, except for Eco52148, showed resistance against gentamicin, amikacin, netilmicin and tobramycin. On the other hand, isolates Eco52148 and Kpn47733 showed resistance to tetracycline (Table 1).

WGS performed on the Sequel I platform and assembly performed on Microbial Assembly pipeline resulted in complete, closed chromosomes and plasmids shown in Table 2. WGS revealed that K. pneumoniae isolates Kn47733, Kpn50595, and Kpn51015 belonged to sequence types (STs) 147, 11 and 15, respectively. All these three STs have been considered as “high risk” clones (Woodford et al., 2011). The E. coli isolate Eco52418 was assigned to ST167. Several studies have reported the association of ST167 E. coli with the dissemination of resistance genes, especially of the carbapenemase-encoding gene bla_NDM–5 (Mani et al., 2017; Sánchez-Benito et al., 2017; Sun et al., 2018; Xu et al., 2019).

Furthermore, analysis of WGS data showed that all four isolates carried different combinations of carbapenemase-encoding genes (Kpn47733: bla_NDM–5 + bla_OXA–181; Kpn50595: bla_NDM–1 + bla_OXA–181; Kpn51015: bla_NDM–1 + bla_OXA–244; Eco52418: bla_NDM–5 + bla_OXA–244) (Table 2). Noteworthy, based on our knowledge, this is the first report of Enterobacterales, carrying bla_OXA–244 gene, isolated from the Czech Republic. Additionally, all isolates exhibited a wide variety of resistance genes conferring resistance to β-lactams, aminoglycosides, sulfonamides, macrolides, lincosamides, streptogramin b, fosfomycin (low-level resistance), fluoroquinolones, chloramphenicol, tetracyclines, and/or rifampicin (Table 2).

The carbapenem resistance phenotypes of all clinical strains were transferred to azide-resistant E. coli A15 by conjugation (Supplementary Table 1). For isolates Kpn47733, Kpn50595 and Kpn51015, all transconjugants carried both carbapenemase-encoding genes, while only the bla_NDM–5 gene was identified in the transconjugants of the E. coli isolate Eco52148.

Analysis of contigs carrying carbapenemase-encoding genes showed that, in isolate Kpn51015, the bla_OXA–244 gene was found on a plasmid (p51015_OXA-244) of 71402 bp in size, while the respective gene was localized in the chromosomal contig of E. coli isolate Eco52418. In both isolates, the bla_OXA–244 gene was bounded by two copies of IS1 insertion sequence in parallel orientation (Supplementary Figure 1), forming a composite transposon, named Tn51098. In the E. coli isolate Eco52418, the Tn51098 transposon was integrated into an open reading frame (ORF) encoding an HNH endonuclease (nts 576701 to 580010 in GenBank accession no. CP050382), as described previously (Potron et al., 2016; Hoyos-Mallecot et al., 2017). Direct repeats of 9 bp (TGAATTGCT) were found at the boundaries of the bla_OXA–244-carrying composite transposon, suggesting its transposition into the E. coli chromosome. However, unlike the isolate Eco52418, an ORF encoding a LysR transcriptional regulator wasn’t found between bla_OXA–244 (downstream) and IS1, in plasmid p51015_OXA-244. Plasmid p51015_OXA-244, which belonged to the incompatibility group FII (IncFII), exhibited extensive similarity with IncFII plasmids from E. coli strains D181 and F5176C6 (GenBank accession nos. CP024250 and CP024669, respectively) (Supplementary Figure 2A). Unlike p51015_OXA-244, those plasmids were negative for the presence of bla_OXA–244 gene. Among p51015_OXA-244, no resistance genes other than bla_OXA–244 were identified.

On the other hand, the bla_OXA–181 carbapenemase-encoding gene was identified on a ColKP3 plasmid (p47733_OXA-181) of 6812 bp in size, in isolate Kpn47733, while a bla_OXA–181-carrying plasmid (p50595_OXA-181; 51140 bp), being an IncX3-ColKP3 fusion, was identified in isolate Kpn50595. In both plasmids, the bla_OXA–181 genes were surrounded by identical sequences (Supplementary Figure 3). In comparison with the archetypal ColE2-type plasmid pKP3-A carrying bla_OXA–181 (Potron et al., 2011), p47733_OXA-181 was composed only of repA and bla_OXA–181 genes (Supplementary Figure 2A). The mob genes, encoding proteins that form a plasmid mobilization system, were not found in plasmid p47733_OXA-181. One additional difference between the two plasmids was the presence of Tn5403 transposon in p47733_OXA-181. The Tn5403 transposon has been previously found in bla_NDM–1-positive IncN2 plasmids, like plasmid pJN24NDM1 characterized from a ST405 E. coli from China (Hao et al., 2019). On the other hand, plasmid p50595_OXA-181 was almost identical to plasmids p1-Ec-BERN-042 (100% coverage, 99.99% identity; GenBank accession no. CP042935) and pOXA181_29144 (100% coverage, 99.99% identity) (Supplementary Figure 2A). Plasmid pOXA181_29144 was previously characterized from a ST18 K. pneumoniae strain (Kpn-29144) isolated, in 2015, in the Czech Republic (Skalova et al., 2017). Similar to pOXA181_29144, which was transferable by conjugation (Skalova et al., 2017), a complete tra locus was found in the sequence of p50595_OXA-181. Also, the qnrS1 gene, conferring low-level resistance to fluoroquinolones, was identified in the sequence of p50595_OXA-181.

The bla_NDM–1 carbapenemase-encoding gene was found on two different plasmid types (Supplementary Figure 4). In isolate Kpn51015, a bla_NDM–1-positive plasmid (p51015_NDM-1) of 353810 bp in size was identified, while a 193462-bp plasmid (p50595_NDM-1) carrying bla_NDM–1 was found in isolate Kpn50595. Plasmid p51015_NDM-1 exhibited extensive similarity to bla_NDM–5-carrying plasmid pKpvST383L (99% coverage, 99% identity; GenBank accession no. CP034201) (Supplementary Figure 2B), characterized from a ST383 K. pneumoniae recovered in London. p51015_NDM-1 belonged to a novel IncH plasmid group, harboring FIB and HIB replicons previously observed in bla_NDM–1-carrying plasmid pNDM-MAR characterized from a ST15 K. pneumoniae isolated in Morocco (Villa et al., 2012). In the MDR of p51015_NDM-1, beside bla_NDM–1, aph(3’)-la, aph(3’)-VI, armA, mph(A), mph(E), msr(E), qnrS1, sul1, sul2, and dfrA5 genes, conferring resistance to aminoglycosides, macrolides, streptogramin b, fluoroquinolones, sulfonamides and trimethoprim, were found (Figure 1). Additionally, plasmid p51015_NDM-1 carried tellurium resistance genes (terZABCDEF), commonly associated with this plasmid family (Zingali et al., 2020). Unlike p51015_NDM-1, p50595_NDM-1 was an IncF_K1-FIB plasmid being a fusion derivative of previously characterized plasmids, like pUCLAOXA232-3, p51015_CTX_M_15, pKPX-1, and pGR-1870 (GenBank accession no. CP012564, CP050379, AP012055 and KF874498) (Supplementary Figure 2B). In p50595_NDM-1, the bla_NDM–1 gene was found in a genetically distant MDR region than observed in plasmid p51015_NDM-1. The MDR region of p50595_NDM-1 also contained bla_CTX–M–15, aac(3)-IIa, aacA4 (2 copies), rmtF (2 copies), catB (2 copies) and arr-2 (2 copies) genes conferring resistance to β-lactams, aminoglycosides, chloramphenicol and rifampicin (Figure 1). Additionally, plasmid backbone of p50595_NDM-1 included an arsenate resistance region. In both bla_NDM–1-carrying plasmids, several insertion sequences (ISs) that could be involved in the organization of their MDR regions were found.

Furthermore, the bla_NDM–5 gene was found in two IncFII plasmids, p47733_NDM-5 and p52418_NDM-5, which contained complete tra operons. The FII (allele 2) plasmid p47733_NDM-5 (103085 bp), found in K. pneumoniae Kpn47733, showed extensive similarity to the bla_NDM–5-carrying plasmid pCRKP-2297_2 (99% coverage, 99.96% identity; GenBank CP024836) (Supplementary Figure 2B) characterized from K. pneumoniae strain CRKP-2297 recovered in South Korea. In p47733_NDM-5, the bla_NDM–5 gene was found in a MDR region of 31159 bp (nts 38315-69473 in GenBank accession no. CP050367) in size. This MDR region, which was bounded by two copies of IS26 element, also included bla_TEM–1, aadA2 (2 copies), sul1, rmtB, mphA and ermB genes conferring resistance to β-lactams, aminoglycosides, sulfonamides, macrolides, lincosamides, and streptogramin b. An additional resistance gene, dfrA12, being the unique gene cassette of a class 1 integron was identified ∼26 Kb upstream of the p47733_NDM-5 MDR region (Figure 2). While both bla_NDM–5-carrying plasmids belonged to IncFII group, the FII (allele 36) plasmid p52148_NDM-5 (121872 bp), found in E. coli isolate Eco52148, harbored a second replicon, FIA (allele 4) (nts 85531-86286 in GenBank accession no. CP050384). Plasmid p52148_NDM-5 exhibited limited nucleotide similarity (53% coverage, 100% identity) against p47733_NDM-5. However, it was highly similar to bla_NDM–5-carrying plasmid p1ESCUMpO83_CORR (100% coverage, 99.98 identity; GenBank accession no. CP033159) (Supplementary Figure 2B) characterized from a pathogenic E. coli strain in India. In p52148_NDM-5, the bla_NDM–5 gene was found in a MDR region of 28400 bp in size (nts 105368-121872 and 1-11895 in GenBank accession no. CP050384). In both plasmids, p47733_NDM-5 and p52148_NDM-5, the genetic environment of bla_NDM–5 was identical (Figure 2). The aadA2, dfrA12, sul1, mphA and tetA resistance genes were also identified in the MDR region of p52148_NDM-5. The MDR region of p52418_NDM-5 was inserted downstream the pemIK operon, as it was also observed in p47733_NDM-5.

Finally, in all four carbapenemase-producing isolates, a diverse number of additional replicons were identified (Table 2). Some of these replicons were associated with important resistance determinants, like bla_CTX–M–15, arr-2 and ermB.

Previous studies have reported the spread of bla_OXA–48-like and bla_NDM-like genes in Enterobacterales recovered from Czech hospitals (Skalova et al., 2017; Paskova et al., 2018). However, based on our knowledge, this study reports the first description of OXA-244-producing Enterobacterales isolated from Czech hospitals. Both OXA-244-producing isolates, also expressed NDM-1 or NDM-5 MβLs. Additionally, two K. pneumoniae isolates, co-producing OXA-181 and NDM-type carbapenemases, were identified. Overall, horizontal gene transfer is the main mechanism involved in the dissemination of bla_NDM-like and bla_OXA–48-like genes, but certain clones have been associated with the spread of these resistance genes (Pitout et al., 2019). Carbapenemase-producing K. pneumoniae and E. coli isolates, characterized during this study, belonged to clones (STs 11, 15, and 147 in K. pneumoniae, and ST167 in E. coli; Table 2), which have been previously characterized as “high risk clones,” and have been reported in association with these carbapenem-resistance mechanisms (Woodford et al., 2011; Pitout et al., 2019). In all cases, carbapenemase-producers were recovered from patients with travel history or previous hospitalization abroad. The endemicity of OXA-181 and NDM-5 carbapenemases among Enterobacterales isolated in the Indian subcontinent has been reported in several studies (Lascols et al., 2013; Krishnaraju et al., 2015; Ahmad et al., 2019; Pitout et al., 2019). Interestingly, a recent report from the United States described the characterization of a ST147 K. pneumoniae harboring bla_NDM–5 and bla_OXA–181 from a patient, who was previously hospitalized in India (Rojas et al., 2017). Also, North African countries, and especially Egypt, represent a geographical region, where bla_OXA–48-like and/or bla_NDM-like genes are highly disseminated among Enterobacterales (Tafoukt et al., 2017; Soliman et al., 2020a,b). Studies from Italy have described the import of NDM-1-producing K. pneumoniae isolates from Egypt (Principe et al., 2016; Nucleo et al., 2020). Additionally, the import of OXA-244-producing E. coli isolates from countries in Northern Africa was observed in a surveillance from Denmark (Hammerum et al., 2020). Finally, two studies have documented the spread of NDM-1-producing K. pneumoniae isolates in Mauritius (Poirel et al., 2012; Holman et al., 2017), speculating a link with India, due to the geographical and cultural links between the two countries. In 2018, another study described the characterization of a K. pneumoniae isolate co-producing NDM-1 and OXA-181 carbapenemases, recovered from a patient, who had previously went to Mauritius (Allyn et al., 2018). These findings highly underline that import of carbapenemase-producing isolates via travel or/and hospitalization abroad could represent a risk for a further dissemination of these isolates in Czech hospitals. However, epidemiological data don’t confirm the scenario regarding the spread of Enterobacterales co-producing NDM- and OXA-48-like carbapenemases in Czech hospitals (Hrabak, unpublished results).

Although the main limitation of this study was the small number of Enterobacterales isolates co-producing OXA- and NDM-type carbapenemases, which were collected in clinical microbiology laboratories, analysis of WGS data revealed that all four isolates harbored a huge variety of genes conferring resistance to several categories of antibiotics. Additionally, inspection of contigs carrying carbapenemase-encoding genes showed that different genetic structures and replicon types were involved in the dissemination of these resistance determinants. These contigs also included a huge variety of insertion sequences that might be involved in the organization of MDR regions, conferring resistance to several antibiotic categories thus, limiting therapeutic options. Thus, in addition to different combinations of carbapenemase-encoding genes, the variability of replicon types, genetic structures, resistance genes and mobile elements observed among the studied isolates indicated the genetic plurality involved in the acquisition and dissemination of determinants encoding OXA/NDM carbapenemases.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ncbi.nlm.nih.gov/genbank/, CP050360–CP050384.

CP and IB played an important role in interpreting the results and in writing the manuscript. KC, LK, and JH helped to acquire data. KC, LK, VM, and IB carried out experimental work. CP supervised the experiments and revised the final manuscript, which was approved by all authors.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.