Enterobacterales are the most prevalent opportunistic pathogens in hospital settings, causing sepsis, pneumonia, soft tissue infections, urinary tract infections, and others. To make this problem worse, the emergence of resistance to β-lactams, the most commonly used family of antibiotics, has drastically reduced treatment options for those infections.1 Since Alexander Fleming described the penicillin in 1928, a large number of β-lactams have been developed belonging to five groups: penams, penems, cephems, monobactams, and carbapenems, all of which possess a β-lactam ring. The carbapenems group is the election to treat infections caused by bacteria producing extended-spectrum β-lactamases (ESBLs) or AmpC cephalosporinases.2

In recent years, the carbapenem-resistant Enterobacterales (CRE) have rapidly emerged and disseminated worldwide. This resistance is caused by either production of carbapenemases or a combination of mechanisms such as the presence of different ESBLs (or AmpC) in addition to porin loss. The evidence suggests that clonal dissemination of carbapenemase-producing Enterobacterales (CPE) plays a critical role in hospital outbreaks of life-threatening infections, which have increased mortality, morbidity and hospitalization costs.3 However, mobile genetic elements (MGEs) including plasmids, transposable elements (TEs), and integrons are probably the most important factors implicated with the dissemination of carbapenemase genes among different bacterial species. TE, including insertion sequences (ISs) and transposons (Tns), are discrete DNA structures, located in plasmids or chromosomes, which can move (through the activity of self-encoded transposases) to new locations in the genome. The ISs are the smallest TEs that can move resistance genes as part of a composite transposon; a gene bounded by two copies of the same or related IS that can move as a single unit.4,5 In this review, we describe different MGEs associated with carbapenemase genes in Enterobacterales members reported in South American countries.

A search of literature published in English or Spanish until August 2019 was performed using PubMed, Google Scholar, LILACS and SciELO databases. In a first-round, the keywords ‘mobile genetic elements AND carbapenemase resistance, carbapenemase AND Enterobacteriaceae or Enterobacterales AND (Argentina OR Brazil OR Bolivia OR Colombia OR Chile OR Ecuador OR Guyana OR French Guiana OR Paraguay OR Peru OR Suriname OR Uruguay OR Venezuela)’ were used. Abstracts or conference reports were not included unless the article form was unavailable. All selected articles were subjected to a second search round for selection of specific studies using molecular analysis criteria: (a) description of clonality based in sequence type (ST) using multilocus sequence typing (MLST) or whole genome sequencing and (b) descriptions of MGEs associated with carbapenemase genes in Enterobacterales species, including plasmid incompatibility groups (based on replicon typing or restriction digestion performed with S1 nuclease) and/or description of TE or integrons structures using DNA sequencing.

Although some successful bacterial clones can disseminate resistance vertically (e.g. K. pneumoniae belonging to CC258 carrying Tn4401/blaKPC), carbapenemase genes (as well as other antibiotic resistance determinants) are frequently transmitted horizontally between different enterobacterial clones, species, and genera.40 Furthermore, these genes also disseminate among different plasmids and chromosomes through transposition or recombination events.41

In South America, we observed the same patterns reported worldwide (Table 1), which indicates that some clones are being transmitted among South American countries (or elsewhere), raising the need for tracking carbapenemase clone dissemination. There is also evidence of carbapenemase gene mobility between plasmids and clones which potentially facilitates the emergence of new epidemic CRE clones such as K. pneumoniae ST11 and the hypervirulent clone ST25 (Table 1). This type of mobility is more difficult to track; plasmids carrying these genes could be identified (by replicon typing or plasmid/whole genome sequencing), but the rearrangements due to recombination and transposition sometimes complicate their traceability.8,42 A clear example of this evolution is the blaKPC-2 gene in K. pneumoniae CC258, which has been associated with IncFII derivatives such as pKpQIL43 and other plasmid incompatibility groups in Argentina, Brazil, Colombia, and Ecuador (Table 1) which have been recovered from K. pneumoniae CC258, non-CC258, and other members of the Enterobacterales.44,45 This variability of plasmids carrying similar blaKPC genetic environments suggests active plasmid rearrangements driven by transposons (e.g. Tn4401). The blaKPC-2 dissemination from the original host (K. pneumoniae) to other species of Enterobacterales is consequence of this evolutionary process: an active transposon harboring antimicrobial resistance genes finding broad (e.g. L/M, N, W) or wide host range plasmids (e.g. IncF, IncH), like it was found in South America (Table 1).46

Another approach to partially assess some of these complexities could be analyzing the genetic environment in which these genes are located. A classic example would be Tn4401, a Tn3-type, 10-kb mobile transposon frequently associated with blaKPC genes. Its structure consists of a Tn3 transposase gene (tnpA), a resolvase gene (tnpR) and additional insertion sequences (ISKpn6 and ISKpn7) delimited by two 39-bp imperfect inverted repeats (IRs).47 The isoforms (a to h) are classified based in nucleotide deletions upstream of blaKPC gene (affecting its promotor region and the KPC-expression), lacking genes or both: for instance, isoform “a” has a 99bp deletion; isoform “b” has no deletion; isoform “c”, a 215bp deletion; isoform “d”, a 68bp deletion; isoform “e”, a 255bp deletion; isoform “f”, a truncation in tnpA, absence of tnpR, ISKpn7 left part and Tn4401 IRL-1; isoform “g” is similar to “f” plus a 215bp deletion (like isoform c); and isoform “h” has a 188-bp deletion.48 (Fig. 1). Non-Tn4401 elements are other genetic environments of the blaKPC gene. These are complex DNA structures sharing some Tn4401 elements such as ISKpn6. Some examples are shown in Fig. 1.11,49,50 In South America the isoforms Tn4401a and b were the most commonly found in regional studies (e.g. in Argentina, Colombia, Brazil, Chile, Ecuador, and Venezuela), but NTEs and the isoforms Tn4401c and d were also described in Brazil, respectively, mainly in non-K. pneumoniae isolates (Table 1). As mentioned before, this lateral dissemination between genera and species would be more related to the type of conjugative plasmid where the transposon ‘landed’.

Fig. 1.

Transposable elements associated with blaKPC, blaOXA-like and blaNDM carbapenemases found in the Enterobacterales members.

(0.33MB).

Currently, more than 70 MBLs (chromosomally- and plasmid-encoded) have been reported and grouped based on DNA sequence similarity.51 NDM is one of the most successfully plasmid-disseminated MBL detected in different members of the Enterobacterales around the world and their frequent association with ISAba125 suggests a possible Acinetobacter spp. origin, a bacterium in which this association is very common.52 Tn125, an ISAba125-based composite transposon, was one of the genetic elements described to be involved in blaNDM dissemination, but in Enterobacteriaceae Tn125 it has been interrupted or truncated, generating a variety of different genetic contexts for blaNDM.53K. pneumoniae, E. coli, and Enterobacter spp. are the predominant carriers of blaNDM, with heterogeneous clonal backgrounds and multiple acquisitions of blaNDM genes across bacterial species, also, different replicon types of blaNDM-carrying plasmids in the Enterobacteriaceae were described, being the most common IncA/C, FIA, FIB, FII and X353 (in our bibliographic review we found less frequent descriptions of MBL-producing Enterobacteriaceae). Reported blaNDM-1 genes were associated with ISAba125 composite transposon Tn125 (Table 1). There are no extensive descriptions of bacterial clones and plasmid incompatibility groups related to blaNDM in South America. In Argentina, there was one description of a blaNDM-1-harboring, IncB/O plasmid recovered from P. rettgeri; in Colombia and Ecuador, blaNDM-1 was detected on IncA/C plasmids from E. coli (Colombia) and K. pneumoniae (both countries). More information has to be developed and analyzed to have a clearer picture of blaNDM spread in South America. A similar scenario was found regarding blaVIM (Argentina-Venezuela) and blaIMP (Brazil), both carbapenemase genes associated to class 1 integrons, also described in Pseudomonas aeruginosa, evidencing that class I integrons are efficient genetic platforms to incorporate MBLs genes and disseminated once a transposon or plasmid is involved.54,55

The OXA-enzymes have higher hydrolysis rates against cloxacillin and oxacillin than other β-lactams. In addition, they are poorly inhibited by clavulanic acid, tazobactam, and aztreonam, and some can inactivate carbapenems.56 We found few descriptions of MGEs associated with these carbapenemase genes in South America; in Ecuador, the blaOXA-48-like gene was related to Tn1999-like.57 However, in Brazil blaOXA-347, a variant of blaOXA-48, showed association with truncated Tn3 and Tn4 transposons.

Current evidence enhances the importance of MGEs for carbapenemase gene dissemination. Except for Argentina, Brazil, and Colombia, the reports of MGEs are scarce in other South American countries. The blaKPC and blaNDM are the most prevalent carbapenemase genes reported in Enterobacterales species are associated with Tn4401/non-Tn4401 elements and ISAba125/Tn125 respectively while blaVIM and blaIMP carbapenemase genes are related to class 1 integrons. The location of transposons or integrons in different plasmid incompatibility groups and bacterial clones denotes their capacity in transferability and mobilization. It is possible that intercontinental dissemination of CRE clones was followed by horizontal gene transfer of plasmids carrying carbapenem resistance genes to local bacteria. Any control measure intended to tackle antibiotic resistance genes dissemination requires the identification of the potential sources of these genes, which nowadays is carried out by DNA polymorphisms. Unfortunately, carbapenem resistance genes have very little polymorphisms. We argue that the study of MGEs associated with these genes could provide very valuable epidemiological information to detect the potential sources.

This study was funded by Instituto de microbiología, Universidad San Francisco de Quito.

The authors declare no conflicts of interest.