Introduction

VIM hydrolyzes nearly all β-lactam except for aztreonam and currently consists of 76 variants (https://www.ncbi.nlm.nih.gov/pathogens/refgene) [1]. The blaVIM genes are mostly found in Pseudomonas and Enterobacteriaceae and occur frequently within class 1 integrons with distinct gene cassette arrays (GCAs). The blaVIM-carrying integrons are always presented in various mobile genetic elements (MGEs) such as plasmids, integrative and conjugative elements (ICEs), and integrative and mobilizable elements (IMEs), enhancing the mobility and dissemination of blaVIM genes [2].There are a wealth of sequenced blaVIM-carrying plasmids, which can be assigned to various incompatibility (Inc) groups such as IncA, IncC, IncN, IncP and IncL/M [3,4,5,6,7]. IMEs/ICEs are frequently located in bacterial chromosomes: ICEs encode self-conjugation modules and thus are able to transfer between cells [8], whereas IMEs needs a helper conjugative element to achieve intercellular mobility [9].

Our previous studies showed the detailed genetic characteristics of two novel carbapenemase-encoding MGEs: SIM-encoding plasmid pHN39-SIM [10] and blaVIM-4-containing ICE Tn6413 [7]. This follow-up study presented the complete sequences of five novel blaVIM-2/-4-carrying MGEs, including two plasmids pJ20133-VIM and p716811-VIM, and two chromosomal IMEs Tn6917 and Tn6918, and one chromosomal ICE Tn6953 from Pseudomonas isolates. These two plasmids were assigned to a novel Inc group IncpSTY. Comprehensive sequence comparisons were then applied to three representative IncpSTY plasmids and 12 chromosome-borne IMEs and ICEs (including the above five ones), providing a deeper understanding of the genetic diversification and evolution of VIM-encoding MGEs in Pseudomonas.

Materials and methods

Bacterial strains and identification

Five clinical isolates P. aeruginosa SE5443, P. putida 716811 and 159349, and P. monteilii J20133 and 918607 were collected from four different Chinese public hospitals (Additional file 1: Table S1). The 16S rRNA genes and the blaVIM genes were detected as described previously [7].

Genomic DNA extraction, sequencing, and sequence assembly

Bacterial genomic DNA was isolated using the UltraClean Microbial Kit (Qiagen, NW, Germany), and sequenced from a sheared DNA library with average size of 15 kb (ranged from 10 to 20 kb) on a PacBio RSII sequencer (Pacific Biosciences, CA, USA), as well as a paired-end library with an average insert size of 350 bp (ranged from 150 to 600 kb) on a HiSeq sequencer (Illumina, CA, USA) [11]. The paired-end short Illumina reads were used to correct the long PacBio reads utilizing proovread [12], and then the corrected PacBio reads were assembled de novo utilizing SMARdenovo (https://github.com/ruanjue/smartdenovo).

Sequence annotation and comparison

Genome sequences were annotated by the Rapid Annotation using Subsystem Technology (RAST) [13] combined with BLASTP/BLASTX/BLASTN [14], Domain (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi), and RefSeq database [15]. Annotation of resistance genes, MGEs, and other features were carried out using the online databases including CARD [16], ResFinder [17], ISfinder [18], INTEGRALL [19], and Tn Number Registry [20]. Multiple and pairwise sequence comparisons were performed using BLASTN. Gene organization diagrams were drawn using Inkscape 1.0 (https://inkscape.org/en/).

Phylogenetic analysis

Indicated nucleotide sequences were aligned using Clustal Omega 1.2.2 [21], and then maximum-likelihood phylogenetic trees were constructed from aligned sequences using MEGA X 10.1.8 [22] with a bootstrap iteration of 1000.

Conjugal transfer

Each indicated blaVIM-carrying MGE was transformed from its wild-type isolate into rifampin-resistant P. aeruginosa ATCC 27853 or PAO1, through conjugal transfer or electroporation experiments. Three milliliters of overnight cultures of each donor and recipient bacteria were mixed together, harvested, and resuspended in 80 mL of Brain Heart Infusion (BHI) broth (BD Biosciences). The mixture was spotted on a 1 cm2 hydrophilic nylon membrane filter with a 0.45 µm pore size (Millipore) that was placed on BHI agar (BD Biosciences) plate and then incubated for mating at 26 °C or 37 °C for 12 h to 18 h. Bacteria were washed from filter membrane and spotted on BHI plates, for selecting a blaVIM-carrying transconjugant. 1500 mg/mL rifampin (for ATCC 27853 or PAO1), together with 4 mg/mL meropenem (for blaVIM) was used as transconjugant selection.

Phenotypic assays and multi-locus sequence typing (MLST)

Bacterial antimicrobial susceptibility was tested by BioMérieux VITEK 2, and interpreted as per the 2020 Clinical and Laboratory Standards Institute (CLSI) guidelines [23]. Activity of Amber class A/B/D carbapenemases was determined with a modified CarbaNP test [24]. The sequence types (STs) of Pseudomonas isolates were identified according to the online Pseudomonas MLST schemes (https://pubmlst.org/organisms).

Results

Identification of three STs for Pseudomonas isolates

Three different STs, namely ST129, ST17, and ST639, were identified from the five Pseudomonas isolates (Additional file 1: Table S1), whose complete genome sequences were determined in this work. P. putida 716811 and 159349 belonged to ST129 and ST17, respectively. P. aeruginosa SE5443 belonged to ST639. ST129 was a novel ST of P. putida. The two P. monteilii isolates J20133 and 918607 could not be assigned with any ST types due to lack of P. monteilii MLST scheme.

Proposal of a novel group of IncpSTY plasmids

Two blaVIM-carrying plasmids pJ20133-VIM and p716811-VIM were identified from the complete genome sequences of the two Pseudomonas isolates J20133 and 716811 (Additional file 1: Table S1). A novel IncpSTY group was proposed from a total of seven available fully sequenced single-replicon plasmids (Additional file 1: Table S2; last accessed August 10th 2020) that included the above two plasmids together with five additional ones from GenBank, because these seven plasmids harbored not only homologous repA (replication initiation protein) genes together with its iterons (Fig. 1) but also similar backbone gene organizations (Fig. 2). Since the Inc group of plasmids could be divided according to the homology of repA genes [25,26,27], a phylogenetic tree (Fig. 1) was constructed based on the repA sequences of these seven plasmids, showing that these seven plasmids could be divided into three separately clustering subgroups IncpSTY-1/2/3. As shown by pairwise comparison of repA sequences, plasmids within each subgroup showed ≥ 95% identity, while those from different subgroups displayed ≤ 95% but ≥ 82% identity (Additional file 1: Table S3). pSTY [28], p716811-VIM, and pHN39-SIM [10] were the first sequenced plasmids of IncpSTY-1, IncpSTY-2, and IncpSTY-3, respectively, and thus were identified as the references of the corresponding IncpSTY subgroups.

Fig. 1
figure 1

Evolutionary relationships of the seven IncpSTY plasmids. a A maximum likelihood phylogenetic tree is constructed from aligned repA sequences. IncP-1α plasmid RK2 [60] is used as the outgroup. Degree of support (percentage) for each cluster of associated taxa, as determined by bootstrap analysis, is shown next to each branch. Bar corresponds to scale of sequence divergence. Triangle indicates the reference plasmid of each IncpSTY subgroup, whilst squares denote the plasmids fully sequenced in this study. b A heatmap of pairwise comparison of repA sequences. See Table S3 for original BLAST coverage and nucleotide identity values

Full size image
Fig. 2
figure 2

Linear comparison of the three IncpSTY plasmids pSTY, pJ20133-VIM, and p716811-VIM. Genes are denoted by arrows. Genes, mobile elements and other features are colored based on function classification. Light blue shading regions denote nucleotide identity > 80%. Red characters represent accessory modules. Numbers in brackets indicate nucleotide positions within corresponding plasmids

Full size image

Comparison of three IncpSTY plasmids pSTY, pJ20133-VIM and p716811-VIM

A detailed comparative genomics analysis was applied to three representative IncpSTY plasmids including the IncpSTY-1 plasmids pSTY and pJ20133-VIM and the IncpSTY-2 plasmid p716811-VIM. The IncpSTY-3 plasmid pHN39-SIM was not included herein because it had been detailed characterized in our previous study [10]. These three characterized plasmids varied in size from about 243 to 321 kb with predicted ORFs from 255 to 282 (Table 1). The modular structure of each plasmid was divided into the backbone, and the accessory modules that were acquired DNA regions and inserted at different sites of the backbone (Fig. 2). pSTY, pJ20133-VIM, and p716811-VIM had > 80% nucleotide identity across > 66% of their backbone sequences, while pSTY and pJ20133-VIM shared > 74% of their backbones with > 95% nucleotide identity (Additional file 1: Table S4). These three plasmids shared the key IncpSTY backbone markers: repA together with its iterons, parABC for partition, and multiple conjugal transfer genes (Fig. 2). A 47.9-kb backbone region orf426–to–orf375 was presented in pSTY, but it was deleted from pJ20133-VIM and p716811-VIM due to insertion of an 11.5-kb Tn7-family transposon remnant (see below) and a 21.6-kb orf282–to–orf237 region, respectively, at the same backbone site. The orf282 − to − orf237 region carried IS1491, mvaT (DNA-binding domain of bacterial xenogeneic silencer MvaT), five unnamed genes (encoding nucleoid-associated protein, predicted NTPase, AAA family ATPase, endonuclease, and putative transposase, respectively), and eight genes encoding hypothetical proteins.

Table 1 Major features of the three IncpSTY plasmids analyzed
Full size table

pSTY, pJ20133-VIM, and p716811-VIM had totally different profiles of accessory modules (Table 1 and Fig. 2): (i) a 59.7-kb paa (phenylacetic acid degradation) region carrying paa locus and mer (mercury resistance) locus, Tn6734, Tn6603a, and ISPsp16 in pSTY; (ii) a 67.2-kb MDR (multi-drug resistance) region, the 11.5-kb Tn7-family transposon remnant, IS2000:P.ae.I3-like intron, ISPmo2, ISPpu22, ISPmo3, and ISPmo5 in pJ20133-VIM; and (iii) a 29.6-kb MDR region, Tn6735, orf282–to–orf237 region, IS1491, and ISPsy2 in p716811-VIM.

The integration of the 59.7-kb paa region into pSTY led to truncation of orf246–to–orf846 region (as originally observed in pJ20133-VIM) into a 138-bp orf846 remnant (Fig. 2). A presumed prototype paa region was terminally bracketed by two copies of Tn5563 [29]; the right copy was intact while the left copy was truncated due to insertion of Tn6481 in the 59.7-kb paa region from pSTY (Additional file 1: Fig. S1). This 59.7-kb paa region still contained a complete set of phenylacetic acid degradation genes (Additional file 1: Fig. S1). Tn6734 was an IME containing a phenol degradation gene locus (Additional file 1: Fig. S2a). Tn6603a was a Tn3-family unit transposon containing an oxidative stress defense gene osmC (Additional file 1: Fig. S2b). The 67.2-kb MDR region from pJ20133-VIM contained In58, Tn5046 and ΔTn512, a 3.6-kb Tn4662a remnant, and several IS elements (Additional file 1: Fig. S3).

According to the previous grouping scheme of Tn7-family transposons [30], the phylogenetic tree was constructed based on the nucleotide sequences of tnsA, which encoded the endonuclease responsible for excision of Tn7-family transposons. A phylogenetic analysis of Tn6735, the 11.5-kb Tn7-family transposon remnant, Tn6922, and Tn6921, together with additional 12 representative sequenced Tn7-family transposons from GenBank based on the tnsA genes, indicated that the first two belonged to a novel Tn6735 subfamily and an unnamable subfamily (it had no intact sequenced transposon), respectively (Fig. 3). Similar to Tn7 [31], both Tn6735 (carrying pyrimidine biosynthesis genes) and the 11.5-kb Tn7-family transposon remnant encoded the core transposition determinants (endonuclease TnsA, transposase TnsB, regulator TnsC, target-site selection protein TnsD, and TnsB-binding sites), and the later one still encoded an additional target-site selection protein TnsE (Additional file 1: Fig. S4).

Fig. 3
figure 3

Evolutionary relationships of the Tn7-family transposons. A maximum likelihood phylogenetic tree was constructed from aligned tnsA sequences. Degree of support (percentage) for each cluster of associated taxa, as determined by bootstrap analysis, is shown next to each branch. Bar corresponds to scale of sequence divergence. Squares denote Tn7-family transposons or transposon remnants sequenced in this study

Full size image

The 29.6-kb MDR region from p716811-VIM has a primary structure of a P.p.I3-like intron with interruption by Tn6740 that carried blaVIM-2, aacA4, qnrVC1, and mer locus (Additional file 1: Fig. S5). As a derivant of the IS66-family transposon Tn5042 [32], Tn6740 retained the whole Tn5042 sequence with integration of a 21.2-kb region that comprised mer-containing Tn3-family unit transposon Tn5563 [29], qnrVC1, and blaVIM-2-carrying In1722 (see below) (Additional file 1: Fig. S5).

Collection of 12 chromosome-borne IMEs/ICEs for sequence comparison

Three blaVIM-carrying chromosomal MGEs Tn6917, Tn6918 and Tn6953 were identified from the complete genome sequences of the three Pseudomonas isolates 918607, 159349, and SE5443 (Additional file 1: Table S1). A detailed sequence comparison was applied to a total of 12 chromosomal MGEs belonging to three groups: two related IMEs Tn6916 and Tn6917; two related IMEs Tn6918 and Tn6919; and eight related ICEs Tn6417, Tn6413, Tn6953, Tn6454, Tn6455, Tn6456, Tn6457 and Tn6458 (Table 2; last accessed August 10th 2020). Tn6917 was the only blaVIM-carrying one belonging to Tn6916-related IMEs in GenBank. Tn6919 was the only one (in addition to Tn6918) belonging to Tn6918-related IMEs in GenBank, and it did not carry blaVIM. The seven Tn6417-related ICEs were selected from GenBank because Tn6417 was used as reference and the rest six ones Tn6413, Tn6954, Tn6955, Tn6956, Tn6957 and Tn6958 carried blaVIM genes. Similar to the above IncpSTY plasmids, the modular structure of each IME/ICE was divided into the backbone and the accessory modules.

Table 2 Major features of the 12 chromosome-borne IMEs and ICEs analyzed
Full size table

Comparison of two related IMEs Tn6916 and Tn6917

Tn6916 (a 51.0-kb prototype IME initially identified in P. monteilii FDAARGOS_171 [33]) and Tn6917 were integrated at the same site downstream of the P. monteilii chromosomal gene dinG (ATP-dependent DNA helicase), and they shared the core backbone markers attL (attachment site at the left end), int (integrase), and attR (attachment site at the right end); in addition, the β-lactam resistance gene ampC was considered as a backbone component of these two IMEs because no associated MGEs were identified for this resistance gene (Fig. 4). Tn6916 thereby had no accessory modules, but blaVIM-2-carrying In528 (see below) was integrated at a site upstream of eamA (transcription regulator) and identified as the sole accessory module in Tn6917 (Fig. 4).

Fig. 4
figure 4

Comparison of the two-related IMEs Tn6916 and Tn6917. Genes are denoted by arrows. Genes, mobile elements and other features are colored based on their functional classification. Shading regions denotes of nucleotide identity ≥ 95%. Numbers in brackets indicate nucleotide positions within the chromosomes of strains FDAARGOS_171 and 918607, respectively. The accession number of Tn6916 [33] for reference is CP014062

Full size image

Comparison of two related IMEs Tn6918 and Tn6919

Tn6918 (a 70.49-kb prototype IME initially found in P. putida 159,349) and Tn6919 were integrated at a site downstream of the P. putida chromosomal gene tRNAThr, and they shared the core backbone markers attL, int1, int2, and attR (Fig. 5). There were at least three major modular differences in the backbone of Tn6919 relative to Tn6918: i) inversion of orf1965, ii) interruption of orf549 and orf2007 due to insertion of ΔISPpu13 and ISPen2, respectively, and iii) replacement of 3’-termimal 22.8-kb orf270–to–orf213 region by 3.4-kb orf255–to–orf282 region (Fig. 5). The ISPa122–mer region was integrated at the same site of the backbones of Tn6918 and Tn6919 (Fig. 5), and it was composed of ISPa122 and a mer-carrying ΔTn5041-like element (Additional file 1: Fig. S6). Tn6918 and Tn6919 acquired blaVIM-2-carrying In1770 (see below) and strAB region, respectively, which were inserted at the same site upstream of orf555 (Fig. 5). The strAB region contained a cryptic unit transposon Tn6920, a strAB-carrying unit transposon Tn6921, a 3.6-kb Tn4662a remnant, and several IS elements (Additional file 1: Fig. S7). Tn6920 and Tn6921 had the core transposition tniABQ− res −tniR modules with the highest sequence identities to Tn5053 [34] and Tn5058a [32], respectively, and thus they belonged to the Tn5053 subfamily of Tn7 family (Fig. 3).

Fig. 5
figure 5

Comparison of the two-related IMEs Tn6918 and Tn6919. Genes are denoted by arrows. Genes, mobile elements and other features are colored based on their functional classification. Shading regions denotes of nucleotide identity ≥ 95%. Numbers in brackets indicate nucleotide positions within the chromosomes of strains 159349 and IEC33019, respectively. The accession number of Tn6918 for reference is CP045553

Full size image

Comparison of eight Tn6417-related ICEs

Tn6417 was a 108.2-kb reference ICE [7] initially found in P. aeruginosa DHS01 [35]. The eight related ICEs Tn6417, Tn6413, Tn6953, Tn6954 (GZAF3_GI) [36], Tn6955 (ICE6440) [3], Tn6956 (ICE6441) [3], Tn6957, and Tn6958 [37] shared the core backbone markers attL/attR (these sequences showed somewhat differences but shared a core motif ‘CCTTCGCCCGCTCCA’), int, cpl (coupling protein), rlx (relaxase), and a F (TivF)-type type IV secretion system gene set (mating pair formation). However, the backbones of these eight ICEs varied in size from 67.8 to 95.1 kb and had at least four major modular differences: (i) presence of unique orf672, orf306, piL1–to–orf381 region, and orf3336–to–orf2514 region in Tn6953; (ii) presence of two highly similar regions lysR2–to–orf1068 and orf348–to–orf1188 in Tn6417 and Tn6413, respectively; (iii) absence of smc–to–orf3006 region from Tn6955; and (iv) interruption of orf2280 owing to insertion of intron E.c.I5 in Tn6954 and Tn6955 (Fig. 6). The first six ICEs were integrated into the P. aeruginosa or A. faecalis chromosomal tRNAGly gene, while the last two ones into the P. aeruginosa chromosomal gene mhpC (methyl ester carboxylesterase) and orf1203 (putative DNA-binding protein), respectively.

Fig. 6
figure 6

Comparison of Tn6417 and the seven related ICEs. Genes are denoted by arrows. Genes, mobile elements and other features are colored based on their functional classification. Shading regions denote nucleotide identity ≥ 95%. Accession number of Tn6417 [7] for reference is EU696790

Full size image

Each ICE carried a single accessory module, and thus there were the resulting eight different accessory modules blaVIM-2-carrying In1779 (see below), Tn6532, Tn6403, Tn6959, a 15.2-kb Tn6346-related region, Tn6960, Tn6961, and Tn6962 from the eight ICEs; the first one was inserted into ftsk (cell division protein), while all the other seven ones were integrated at a site upstream of orf582 and identified as the derivatives of Tn6346 (Fig. 7). Tn6346 was a prototype Tn3-family unit transposon originally identified in Achromobacter spp. AO22 [38] and manifested as a hybrid of the core transposition module tnpAR–res from Tn5051 and the mer region from Tn501 (Fig. 7). The seven Tn6346 derivatives differed from Tn6346 in two major aspects: (i) insertion of IS1071 at the same position within tnpA in all these seven Tn6346 derivatives, and furthermore insertion of ISPa91 at another position within tnpA in only Tn6960; and (ii) insertion of different concise class 1 integrons In159, In127, In1365, In56, In1879, In1853, and In58 into urf2 in these seven Tn6346 derivatives, sometimes leading to truncation of surrounding regions (Fig. 7). The blaVIM genes were found the last five integrons (see below) rather than the first two (Additional file 1: Fig. S8).

Fig. 7
figure 7

Comparison of Tn6346 and the seven related elements. Genes are denoted by arrows. Genes, mobile elements and other features are colored based on their functional classification. Shading regions denote nucleotide identity (light blue: ≥95%; and light pink: ˂95% but ≥ 79%). Numbers in brackets indicate nucleotide positions within the chromosomes of strains DHS01, 6762, GZAF3, FFUP_PS_CB5, CCUG 51971, MUB14, and HSV3483, respectively. The accession number of Tn6346 [38] for reference is EU696790

Full size image

Comparison of eleven bla VIM-carrying integrons

The blaVIM genes were found in the two plasmids pJ20133-VIM and p716811-VIM (Table 1) and the nine chromosomal IMEs/ICEs Tn6917, Tn6918, Tn6413, Tn6953, Tn6454, Tn6455, Tn6456, Tn6457 and Tn6458 (Table 2). The 11 corresponding local blaVIM genetic environments manifested as 11 different class 1 concise integrons, and three of them contained blaVIM-4 while the other nine carried blaVIM-2 (Fig. 8). A complete class 1 concise integron contained IRi (inverted repeat at the integrase end), 5’-conserved segment (5'-CS: intI1–attI1), GCA, 3'-CS (qacED1–sul1–orf5–orf6), tniTn402 (Tn402 core transposition module tniABQ–res–tniR), and IRt (inverted repeat at the tni end). Except for In1443Tn6413 with terminal truncations, all the other tens had the terminal IRi/IRt pairs. Insertions or truncations occurred in 5'-CS or 3'-CS or tniTn402 of all these 11 integrons, especially including In1779 that showed the longest modular structure because of integration of a mer-containing Tn3-family prototype unit transposon Tn6758. In56Tn6596 had the shortest GCA that contained only blaVIM-2, while all the other 10 integrons had the multiple-gene GCAs harboring blaVIM-2 plus one or more additional antibiotic resistance genes.

Fig. 8
figure 8

Comparison of the 11 blaVIM-carrying integrons. Genes are denoted by arrows. Genes, mobile elements and other features are colored based on their functional classification. Shading regions denote of nucleotide identity (light blue and yellow: ≥ 95%; and light pink ˂95% but ≥ 85%). Numbers in brackets indicate nucleotide positions within the corresponding plasmids pJ20133-VIM and p716811-VIM or within the chromosomes of strains FFUP_PS_CB5, 6762, GZAF3, HSV3483, CCUG 51971, MUB14, 159349, 918607, and SE5443

Full size image

Summary of newly identified or designated MGEs

There were eight newly identified MGEs, including two IMEs Tn6917 and Tn6918, one ICE Tn6953, two unit transposons Tn6735 and Tn6740, and three integrons In1770, In1772, and In1779. Additional 13 MGEs (IMEs: Tn6734, Tn6916, Tn6919, and Tn6957; unit transposons: Tn6920, Tn6921, Tn6959, Tn6960, Tn6961, and Tn6962; IS elements: ISPmo2, ISPmo3, ISPmo5) were newly designated but with previously determined sequences. The three previously designated ICEs GZAF3_GI, ICE6440, and ICE6441 were renamed as the standard Tn designations Tn6954, Tn6955, and Tn6956, respectively.

Transferability and antimicrobial susceptibility

pJ20133-VIM, p716811-VIM, and Tn6953, which were selected to represent the IncpSTY-1 and IncpSTY-2 plasmids and the Tn6417-related ICEs, respectively, could be transferred from the relevant wild-type isolates into ATCC 27853 or PAO1 through conjugation, generating the transconjugants PAO1/pJ20133-VIM, PAO1/p716811-VIM, and ATCC 27853/Tn6953 respectively. All these wild-type and transconjugant strains had the Ambler class B carbapenemase activity (data not shown) and were resistant to imipenem and meropenem with minimal inhibit concentration (MIC) values ≥ 16 mg/L (Additional file 1: Table S5), owing to production of VIM enzyme.

Discussion

The present work identifies a novel group of IncpSTY plasmids, which can be further divided into the three subgroups IncpSTY-1/2/3. A detailed genetic dissection analysis is applied to a collection of 15 Pseudomonas MGEs, including three IncpSTY plasmids, two Tn6916-related IMEs, two Tn6918-related IMEs, and eight Tn6417-related ICEs. All these IMEs/ICEs were located within the bacterial chromosomes. At least 22 genes for resistance to seven different categories of antibiotics and heavy metals are identified within these 15 MGEs (Additional file 1: Table S6). Eleven of these 15 MGEs carry the blaVIM genes, which are all located within GCAs of concise class 1 integrons. For 10 of these integrons, the blaVIM genes were presented together with other resistance genes, especially including those for resistance to β-lactams, aminoglycosides, trimethoprim, and rifampicin. These blaVIM-carrying integrons were further integrated into the above plasmids, IMEs and ICEs.

Confirmed by conjugal transfer experiments herein, IncpSTY plasmids and Tn6417-related ICEs have the ability to transfer from one cell to another cell. These two groups MGEs could transfer autonomously by utilizing self-encoded conjugation gene sets [39, 40]. However, the conjugal transfer of Tn6916- and Tn6918-related IMEs are nonautonomous due to lack of essential conjugation genes encoding Cpl and type IV secretion system, and thereby its intercellular transfer is relied on the help of other conjugative elements [40].

blaVIM genes are widely disseminated in Pseudomonas [41,42,43], Enterobacteriaceae [5, 44, 45], Providencia [46, 47], Achromobacter [48], Acinetobacter [49, 50], Aeromonas [51], and Morganella [52]. Currently, at least 183 different blaVIM-carrying integrons [19, 53, 54] have been identified, and they either directly integrate into the chromosomes or reside as the inner-components of various MGEs such as plasmids, unit transposons, ICEs and IMEs [3, 55,56,57]. These results indicate that blaVIM genes have evolved to reside diverse MGEs with the intra- or inter-cellular mobility, which will promote the wide dissemination of blaVIM genes.

Until now, only seven IncpSTY plasmids are identified in six Pseudomonas isolates and one Stenotrophomonas rhizophila isolate, which come from China (n = 5), Japan (n = 1) and Germany (n = 1) (Additional file 1: Table S2). Only six Tn6916-related IMEs and two Tn6918-related IMEs are detected in eight Pseudomonas isolates, two of which are sequenced in this study. The six Tn6916-related IMEs come from China (n = 4), USA (n = 1) and Japan (n = 1), while the two Tn6918-related IMEs from China (n = 1) and Brazil (n = 1). These observations indicate that these three group MGEs are mainly prevalent in Pseudomonas isolates and meanwhile still not widely disseminated in multiple geographic regions. By contrast, Tn6417-related ICEs are frequently identified in Pseudomonas [7, 58, 59] and also occasionally in Bordetella, Alcaligenes, Klebsiella pneumoniae, Achromobacter xylosoxidans, Morganella morganii, Aeromonas caviae, Casimicrobium huifangae, and Delftia acidovorans. Tn6417-related ICEs have been already disseminated worldwide in a plenty of bacterial species.

Conclusion

The four groups of blaVIM-carrying MGEs characterized in this study have a wide range of hosts including Pseudomonas, and they are able to transfer across different bacterial species. Integration of blaVIM genes into these MGEs contributes to the accumulation and distribution of blaVIM genes and enhances the ability of bacteria to survive under selection pressure of carbapenems especially in hospital settings. Moreover, these four groups of MGEs display a high-level diversification in modular structures, which have complex mosaic natures and carry a large number of drug resistance genes (particularly in the MDR regions in these MGEs). This study provides a deeper insight into the genetic diversification and evolution of VIM-encoding MGEs in Pseudomonas.