Understanding the rapid spread of antimicrobial resistance genes mediated by IS26

Abstract Insertion sequences (ISs) promote the transmission of antimicrobial resistance genes (ARGs) across bacterial populations. However, their contributions and dynamics during the transmission of resistance remain unclear. In this study, we selected IS26 as a representative transposable element to decipher the relationship between ISs and ARGs and to investigate their transfer features and transmission trends. We retrieved 2656 translocatable IS 26 ‐bounded units with ARGs (tIS26‐bUs‐ARGs) in complete bacterial genomes from the NCBI RefSeq database. In total, 124 ARGs spanning 12 classes of antibiotics were detected, and the average contribution rate of IS26 to these genes was 41.2%. We found that IS 26 ‐bounded units (IS26‐bUs) mediated extensive ARG dissemination within the bacteria of the Gammaproteobacteria class, showing strong transfer potential between strains, species, and even phyla. The IS26‐bUs expanded in bacterial populations over time, and their temporal expansion trend was significantly correlated with antibiotic usage. This wide dissemination could be due to the nonspecific target site preference of IS26. Finally, we experimentally confirmed that the introduction of a single copy of IS26 could lead to the formation of a composite transposon mediating the transmission of “passenger” genes. These observations extend our knowledge of the IS26 and provide new insights into the mediating role of ISs in the dissemination of antibiotic resistance.


INTRODUCTION
Bacterial antimicrobial resistance is a serious worldwide public health problem.Antimicrobial resistance genes (ARGs) are usually associated with mobile genetic elements, and the sequencing of bacterial genomes has revealed a large repertoire of resistance determinants located on various transposable elements, including plasmids, insertion sequences (ISs), transposons, integrons, and integrative conjugative elements 1,2 .These mobile genetic elements promote the acquisition and transmission of ARGs, as they play a vital role in facilitating horizontal gene transfer within or between DNA molecules.ISs are generally small mobile elements encoding the enzymes necessary for their transposition; they are capable of repeated insertion into many different sites within a genome 3 .Moreover, ISs are known to mediate the mobilization of ARGs as part of a composite transposon, that is, an element with two bounded copies of the same or related IS moving as a single unit 4 .A recent study also demonstrated that ISs could interact with conjugative plasmids to influence the horizontal transfer of ARGs 5 .Therefore, ISs are considered to contribute significantly to the spread of ARGs.However, we still lack the knowledge 1 State Key Laboratory of Microbial Resources, Institute of Microbiology, Chinese Academy of Sciences, Beijing, China. 2 College of Life Science, University of Chinese Academy of Sciences, Beijing, China.required to evaluate the ARG content, phylogenetic distribution, and temporal dynamics of specific ISs.
IS26, a member of the IS6 family, is frequently detected in a variety of clinically resistant bacterial isolates and has attracted considerable attention due to its contribution to the spread of antibiotic resistance.Upon insertion, IS26 generates short 8-bp directly repeated (DR) sequences of the target DNA flanking the inserted IS26-based element 6 .Transposition, based on either a replicative or conservative mechanism, may lead to the formation of an IS26-bounded composite transposon comprising two IS26 elements flanking a central DNA segment.This element has been widely reported to be closely genetically related to various ARGs.In the year 1982, an IS26-based transposon, Tn2680, was first reported to mediate the kanamycin resistance 7 .Thereafter, IS26-bounded transposons have been increasingly reported on multidrug-resistant plasmids from clinical bacterial isolates in association with multiple ARGs, such as aphA1, tetM, bla CTX-M-55 , bla NDM , and bla KPC 8-11   , suggesting a role of IS26 in the transmission of these resistance genes.Furthermore, our previous study indicated that IS26 could mediate rapid amplification of bla KPC and increase carbapenem resistance 12 , which has a significant clinical impact.
In this study, we selected IS26, a crucial mobile element for the dissemination of antimicrobial resistance, to understand the IS-mediated rapid spread of ARGs.We performed a comprehensive analysis of IS26 and IS26-bounded units (IS26-bUs) in complete bacterial genomes from the NCBI RefSeq database, explored the relationships among IS26, ARGs, and antibiotic consumption, evaluated the potential of horizontal gene transfer of IS26, and analyzed the temporal trend of IS26 dissemination in bacterial populations.This study reveals a significant contribution of the IS26 element to the dissemination of prevalent ARGs in Gammaproteobacteria, proposes reasons for the frequent transfer, and deepens our understanding of the contribution of ISs to the spread of ARGs.

Prevalent ARG types are enriched in tIS26-bUs
To explore the role of IS26-bUs in the dissemination of antibiotic resistance, we established a comprehensive data set of 2656 translocatable IS26-bUs with ARGs (tIS26-bUs-ARGs) (Figure 1A) belonging to 340 clusters (Table S1) based on 12,142 high-quality complete bacterial genomes of NCBI Re-fSeq genome database.
A total of 5439 ARGs belonging to 124 ARG types that confer resistance to 12 classes of antibiotics were detected in the tIS26-bUs.All of the antibiotic classes encompassed broad-spectrum antibiotics widely used in humans and animals.None of the identified ARGs confer resistance to antibiotics that specifically target Gram-positive bacteria, indicating that the tIS26-bU mainly mediated ARG spread in Gram-negative bacteria (Figure 1B).β-lactam is one of the most widely used classes of antibiotics worldwide; accordingly, β-lactam ARG types were detected most frequently (42) in the tIS26-bU.Aminoglycoside resistance genes (31), fluoroquinolone resistance genes (13), and folate pathway antagonist resistance genes (12) were also prominent, while there were <10 ARGs that confer resistance to other antibiotics (such as tetracycline, macrolide, and polymyxin).The contribution rates of tIS26-bUs to the 124 ARG types ranged from 1.6% to 100% (Figure 1B).The average contribution rate was 41.2%, with a standard deviation of 2.6%, highlighting the role of IS26 in the dissemination of antibiotic resistance.Importantly, some variants of the New Delhi metallo-β-lactamase and extended-spectrum β-lactamase genes, such as bla NDM-3 , bla NDM-4 , bla NDM-9 , bla SHV-5 , bla SHV-44 , bla TEM-216 , bla OXA-534 , and bla VIM-27 , which have emerged as severe threats, were located only on tIS26-bUs, indicating that IS26 is a unique shuttle bus for these ARGs.Specifically, tIS26-bUs commonly carry 1-10 ARGs (average of 2.05).

tIS26-bUs mediate extensive ARG dissemination within the Gammaproteobacteria
To understand the extent to which tIS26-bUs mediate the dissemination of ARGs, we constructed a phylogenetic tree to analyze the phylogenetic relationships among the known hosts of this element (Figure 1C).Overall, 2651 tIS26-bUs-ARGs were found in 69 species of Proteobacteria, and five were found within four species of the phylum Actinobacteria.The majority (2644, 99.7%) of tIS26-bUs were distributed within 64 species of Gammaproteobacteria; only one and six tIS26-bUs were identified in Alphaproteobacteria and Betaproteobacteria, respectively (Figures 1C and S1).These results indicate that Gammaproteobacteria is the main host of tIS26-bUs-ARGs, and that these units are capable of crossing phylum barriers, even between Gram-positive and -negative bacteria.Among all of the species analyzed, four were identified as highly abundant carriers of tIS26-bUs; all these four species are clinically notorious pathogens: Escherichia coli, Klebsiella pneumoniae, Salmonella enterica, and Acinetobacter baumannii (Figure 1C).Furthermore, we found that one strain could carry 1-12 clusters of tIS26-bUs-ARGs and up to 48 ARGs conferring resistance to 10 classes of antibiotic (Figure 1C), suggesting the complexity of resistant bacteria and the high risk of antibiotic resistance mediated by tIS26-bUs-ARGs.Moreover, more than half (581/917 = 63.3%) of the strains carried tIS26-bUs-ARGs only on the plasmids, 26.1% of the strains carried tIS26-bUs-ARGs only on the chromosomes, while a small proportion (10.6%) of the strains carried both plasmid-and chromosomelocated tIS26-bUs-ARGs (Figure 1D).It was interesting that the location of the tIS26-bUs-ARGs showed taxonomy preference among the strains.The tIS26-bUs-ARGs were mainly located on plasmids in strains of Enterobacteriaceae, while they were more likely to be located on chromosomes in strains of non-Enterobacteriaceae, like A. baumannii (Figure 1E).
To evaluate the transfer capacity of the tIS26-bUs, we analyzed the genomic context environments of the 340 clusters.These clusters were inserted into 2-96 different genomic environments (mean of 5.42; standard deviation of 9.76) (Figure 2A) and identified in 1-17 species, 1-10 genera, The fifth ring shows the number of tIS26-bUs-ARGs detected on each strain.The sixth ring shows the number of ARGs carried by each strain.The seventh ring shows the classes of the ARGs carried by each strain.(D) Venn diagram shows the relationship between the strain sets of plasmidlocated tIS26-bUs-ARGs and that of chromosome-located tIS26-bUs-ARGs.(E) The distribution of plasmid-and chromosome-located tIS26-bUs-ARGs of strains from different species.The red stars indicate the species of Enterobacteriaceae.ARGs, antimicrobial resistance genes; tIS26-bUs, translocatable IS26-bounded units; tIS26-bU-ARGs, translocatable IS26-bounded units with antimicrobial resistance genes.
1-5 families, 1-4 orders, 1-2 classes, and 1-2 phyla (Figure 2B).This implies that tIS26-bUs are frequently transferred, including across phylogenetically distant bacteria.The most frequent tIS26-bU cluster is Cluster2630, consisting of an aph(3′)-Ia gene flanked by two IS26s (Figure S2A).A total of 143 tIS26-bUs were identified in Cluster2630; they were located in 96 different genomic environments and distributed among 17 species within Gammaproteobacteria (Table S1); 101 tIS26-bUs were located on plasmids, and 42 were located on chromosomes.Overall, the tIS26-bUs within Cluster2630 demonstrated its high capacity to disseminate among bacteria.Cluster2041, carrying the carbapenem resistance gene bla KPC-2 (Figure S2B), was found in nine different genomic environments in K. pneumoniae, suggesting frequent transfer of this transposon and a high risk of spreading of carbapenem-resistant genes by tIS26-bUs.Cluster2557 carrying an aph(3′)-Ia and a transcriptional regulator (Figure S2C) was found in the Gram-positive bacteria phylum Actinobacteria, as well as in six species, four genera, three families, two orders, and one class of the phylum Proteobacteria, indicating high ability of the transposon to facilitate horizontal gene transfer of the aminoglycoside resistance gene across phylogenetically distant bacteria.

Expansion of tIS26-bUs-ARGs in bacterial populations over time
To understand the temporal dynamics of tIS26-bUs, we calculated the proportion of tIS26-bUs-ARG-positive genomes in sequenced strains over time.The obtained linear regression equation, Y = 0.001103 X -2.162 (R 2 = 0.54, p < 0.0001), indicated a trend toward expansion of tIS26-bUs.From 2005 to 2017, the proportion of genomes with tIS26-bUs-ARGs increased steadily.The number of tIS26-bUs-ARG-positive strains increased from 6 to 169, and the proportion of tIS26-bUs-ARG-positive strains in sequenced strains rose from 3.2% to 12.4% (Figure 3).This result demonstrates the expansion of tIS26-bUs-ARGs among bacterial populations over time.The Cochran-Armitage and the Mann-Kendall tests were performed to determine the temporal trend of the strains with tIS26-bUs-ARGs.The p < 0.0001 convincingly indicated that the proportion of tIS26-bUs-ARG-positive strains within the sequenced strains followed a monotonic increasing trend.We also analyzed the dynamics of tIS26-bUs-ARGs from different sources, including human-, animal-, and environment-derived strains.The scatter plots and Mann-Kendall test results showed that the proportion of tIS26-bUs-ARG-positive strains also followed a monotonic increasing trend across all three sources (Figure S3A and Table S2).We confirmed this dynamic trend by conducting random sampling, which clearly showed that the tIS26-bUs-ARGs disseminated among bacterial populations over time (Figure S3B and Table S2).Additionally, we compared the proportion of tIS26-bUs-ARG-positive strains from different sources.There was no significant difference between the proportions of tIS26-bUs-ARG-positive strains in human-and animal-origin strains, while the proportions of human-origin were significantly higher than those of environmental-origin (Figure S3C), suggesting the higher effect of human and animal than environment onto the tIS26-bUs-ARG-positive strains.
The host of tIS26-bUs-ARGs also changed over time.From 1980 to 1989, tIS26-bUs-ARGs were detected in only five species of Gammaproteobacteria. From 1990 to 1999, tIS26-bUs-ARGs were not only distributed in two additional species  of Gammaproteobacteria, but also in one species of Actinomycetia.From 2000 to 2009, the host range of tIS26-bUs-ARGs increased by seven species of Gammaproteobacteria; and from 2010 to 2017, tIS26-bUs-ARGs spread to additional 51 species of Gammaproteobacteria, four species of Betaproteobacteria, one species of Alphaproteobacteria, and three species of Actinomycetia (Table S3).
The spread of β-lactam resistance mediated by IS26-bUs correlates with antibiotic usage As β-lactam ARG types are most abundant in the IS26-bUs, we performed a correlation analysis of the relationship between tIS26-bUs carrying β-lactam resistance genes (tIS26-bU β-lactam ) and global consumption of cephalosporins and carbapenems between 2000 and 2015, using data from ResistanceMap.The results showed that the temporal trend of the proportion of tIS26-bU β-lactam -positive strains was significantly associated with the consumption of cephalosporins or carbapenems in more than half (9/14 = 64.3%) of the countries (Table S4).The dynamics of the consumption of cephalosporins correlated with the proportion of tIS26-bU β-lactam positive strains in Brazil, China, Germany, India, and the United States, while the consumption of carbapenems correlated with transposon dynamics in Australia, Canada, China, Denmark, India, and South Korea.These results suggest that the rapid spread of ARGs mediated by tIS26-bUs could be promoted by increased consumption of antibiotics.

IS26-bUs exhibit a nonspecific target site preference
To identify the reason for the frequent horizontal gene transfer of IS26-bUs, we analyzed the characteristics of the 8-bp direct repeat (DR) sequences generated by IS26 insertion to determine whether the IS26 elements exhibit a preference for target DNA sequences.A total of 226 8-bp DR sequences of IS26 elements were obtained.We first classified the sequences into clusters using CD-HIT (nucleotide similarity of 100%), and 168 clusters were obtained, indicating the diversity of the DR sequences.We then calculated the mean GC content of the 8-bp DR sequences and found that the value varied dramatically, from 0% to 87.5% (Figure 4A), suggesting that there was no obvious GC content preference of the target DNA sequences of IS26.We also analyzed sequence conservation of the nucleotides of the 8-bp sequences and found a slight bias toward thymine at base positions 2-4 but no significant preference (Figure 4B).The information content of the 226 8-bp DR sequences at any given position in the consensus sequence was lower than 0.1 bit, suggesting low conservation.Moreover, the emergence odds of four bases at any given position appeared random (range: 11.9%-39.9%;mean of 25%; standard deviation of 1%), indicating that the probability of occurrence of the four bases at each position is almost equal.Accordingly, we can conclude that IS26-bUs insert into DNA sequences without preference.

Unexpected rapid reaction to high antibiotic pressure of single IS26
During data analysis, we noticed that there were some strains carrying only one IS26 gene (single-copy) in the genome (located on plasmids or chromosomes).To explore the role of single-copy IS26 element in the dissemination of antimicrobial resistance, we constructed a plasmid (pUC57:IS26-bla KPC ) carrying an IS26 element, an aminoglycoside resistance gene (ampR), and a carbapenem resistance gene (bla KPC ) and introduced it into E. coli MG1655.Test cell cultures were exposed to antibiotic pressure (the concentration of carbapenem gradually increased from 1 to 256 μg/ml, doubled per inoculation), and control cells were cultured in a medium without antibiotics.We subsequently performed nanopore whole-genome sequencing of the entire cell culture that survived under the highest carbapenem concentration (256 μg/ml) to detect the integration mediated by a single IS26 and determine the effect of antibiotic pressure.Notably, the plasmid (pUC57:IS26-bla KPC-2 ) was observed randomly inserted into the chromosome of E. coli MG1655 cells challenged with carbapenem (Figure 5), demonstrating the mediation role of IS26 during the insertion.The plasmid was (A) (B) inserted into two sites of the chromosome and was flanked by two IS26 elements, thereby generating a new IS26-bounded transposon carrying two ARGs, ampR and bla KPC (Figure 5 and Table S5).One insertion site of the composite transposon was within the ybiL gene, resulting in a truncated gene; the other was in a noncoding region between the mdaB and uhpC genes.These results indicate that a single IS26 can prompt random formation of ARG-carried composite transposons, effectively increasing resistant levels under high antibiotic pressure.We also compared the hosts of tIS26-bUs-ARGs to those of IS26 and detected a larger host range for IS26 than for tIS26-bUs-ARGs.In total, 30 species belonging to 11 genera, three families, and two orders carried IS26, but did not contain tIS26-bUs-ARGs (Figure S4); this shows the potential for ARGs to reach new hosts mediated by IS26.

DISCUSSION
It is well known that IS26 mediates the transfer of many ARGs; however, the preference of ARGs carried by IS26, the phylogenetic distribution of the hosts, and the temporal dynamics of IS26 remain elusive.In this study, we aimed to answer these questions and shed light on the dissemination of antibiotic resistance mediated by IS26.First, we demonstrated that 124 ARG types conferring resistance to 12 antibiotic classes were frequently carried by tIS26-bUs.
Other studies have shown that ISAba1 and its transposon were mainly responsible for the movement of the bla OXA carbapenemase gene 13 , and 60 ARG genotypes and nine ARG phenotypes were detected on class 1 integrons 14 .
Compared with these mobile elements, IS26 carried more diverse and greater numbers of ARGs.Consistent with other studied mobile elements, tIS26-bUs were widely distributed in Gram-negative bacteria upon host analysis.In addition, tIS26-bUs-ARGs were predominantly found in clinically associated pathogens of Gammaproteobacteria, which was consistent with previous studies 10,15,16 .Notably, the average contribution rate of IS26 to ARGs was 41.2%, and the contribution rate to some genes even reached 100%.These results suggest that IS26 plays an important role in the spread of ARGs in bacteria, especially clinical pathogens.Bacterial antibiotic resistance is becoming an increasingly serious problem and often occurs due to the acquisition of resistance genes by previously susceptible bacteria 17 .Our study revealed that tIS26-bUs-ARGs-positive strains are widely spreading in bacterial populations and that this trend is increasing over time.This implies that IS26 is closely associated with the development of bacterial antibiotic resistance.Tn4401 was considered the origin of bla KPC gene mobilization 18 ; however, IS26-based transposons, as a novel genetic environment for bla KPC , are increasingly being reported worldwide, especially in China 19,20 , suggesting a trend for IS26 to become a major genetic element for the acquisition and widespread dissemination of the bla KPC gene.Moreover, the temporal relationship of tIS26-bUs-ARGs-positive strains was closely related to the consumption of cephalosporins or carbapenems, indicating that IS26 can respond to antibiotic pressure.Therefore, antibiotic usage should be strictly regulated and controlled to reduce the spread of antibiotic resistance.
Generally, the insertion of ISs shows target site preferences: for example, IS1 prefers AT-rich regions for insertion 21 , while IS21, IS30, and IS911 all insert close to sequences that resemble their own inverted repeat (IR) [22][23][24] .However, the target sits of IS26 was found with no obvious coincidence in a comparison of 14 insertion sites 25 .In the current analysis based on all the 2656 publicly available tIS26-bUs, we further confirmed that IS26 displayed a preference for random insertion, and thus has more opportunities for insertion.Moreover, previous research demonstrated the movement of the translocatable units formed by a single IS26 26 .Consistent with this, our study also found that the introduction of a single copy of IS26 can lead to the formation of IS26-bounded transposons and the movement of flanking genes to rapidly respond to high antibiotic pressure.Notably, in our previous study, the transfer of the tIS26-bU-bla KPC-2 among two clinical E. coli strains had been observed 10 .These observations illustrate the strong transfer potential of IS26 during the mediation of antibiotic resistance and demonstrate the high risk of tIS26-bU in horizontal gene transfer and the spread of ARGs.
Even though the analyzed genomes retrieved from the NCBI RefSeq database do not represent the entire range of genomes, they served as appropriate representatives of the microbes of interest for the current study.Using IS26 as a mobile element model, we shed light on the relationship between ISs and ARGs and firmly established the critical role of ISs in the dissemination of antibiotic resistance.Given the significant contribution of IS26 to the mediation of ARGs, especially in clinical pathogens, and the relationship observed with antibiotic usage, we strongly suggest that IS26 be used as an indicator for monitoring antibiotic resistance in clinical practice.

Bacterial genome preparation
A total of 12,142 complete bacterial genomes from the NCBI RefSeq genome database (https://ftp.ncbi.nlm.nih.gov/genomes/refseq/bacteria/) were downloaded on March 11, 2021.The isolation time, isolation source, country or region, and taxonomic lineages of the genome's host were retrieved from their GenBank files.The isolation, sequencing, submission, and data releasing of a strain would last at least 3-5 years.Thus, there is a latency between bacterial genome submission and strain isolation.According to the isolation time, we noticed that for the 12,142 downloaded genomes, there are too few genomes isolated after the year 2018.Therefore, we only used strains isolated before 2018 to ensure a complete data set for the analysis of temporal trends of IS26, avoiding sample bias.The isolation source was further divided into human, animal, environmental, other sources, and unknown sources.These data are summarized in Table S3.

Data collection for antibiotic use
Data indicating the time when specific antibiotics received FDA approval for human application were retrieved from the study of Durand 27 .Consumption data for cephalosporins and carbapenems of the countries and regions involved in our study were collected from ResistanceMap (https:// resistancemap.cddep.org/AntibioticResistance.php), as previously described 28 .

Extraction and annotation of IS26-bounded transposons and related sequences
The nucleotide sequence of the IS26 element was downloaded from the ISfinder database 29 .A BLASTN search for the IS26 sequence in complete bacterial genomes within the NCBI database was performed with cut-offs of at least 95% nucleotide similarity and 95% IS26 sequence coverage to annotate the IS26 element.Based on the genomic position of pairwise alignment of IS26 in replicons identified by BLASTN, DNA sequences flanked by two copies of IS26, defined as the IS26-bUs, were extracted.Then, we assigned the IS26-bU sequences to different clusters using CD-HIT 30 , according to a nucleotide similarity of 99%.Next, we extracted the 5-kb upstream and downstream genomic sequences surrounding the IS26-bU and again clustered these sequences using CD-HIT at a nucleotide similarity of 99%.We checked and classified the IS26-bU located in different genomic environments as translocatable, defined as the translocatable tIS26-bU.Moreover, the ARGs were annotated by using ResFinder v4.1.0 31for these tIS26-bUs to obtain the tIS26-bUs-ARGs.The contribution of tIS26-bU to ARGs was calculated by dividing the number of the ARGs on the tIS26-bU by the number of ARGs annotated in the NCBI complete genome, multiplied by 100, as shown in Equation (1).

Construction of phylogenetic tree
The core genes of the tIS26-bUs-ARGs-positive strains were constructed through the Roary v3.13.3 32 based on the protein annotation files of all the strains.Then, the concatenated sequence of the core genes was used to construct the phylogenetic tree by using RAxML 33 .The tree file was further illustrated on the iTol (https://itol.embl.de/itol.cgi).

Statistical analysis
The proportion of tIS26-bUs-ARGs over time was normalized by dividing the genome number of tIS26-bUs-ARGs in a given year by the total number of genomes in the database for that year.Then, linear regression analysis to determine the changing trend of proportion of tIS26-bUs-ARGs-positive strains along with the isolation time thereof was performed using GraphPad Prism software for Windows (ver.8.0.1;GraphPad Software Inc.).For trend analysis, the Cochran-Armitage and Mann-Kendall tests were performed using the "DescTools" (ver.0.99.44) and "trend" (ver.1.1.4)packages in R software (ver.4.1.2).The scipy module (ver.1.5.4) in Python (ver.3.6.13)was used to calculate the Spearman's correlation coefficient.

Target site sequence analysis
The 8-bp DR sequences located on each side of the IS26 elements were extracted from the genomes.Then, the obtained 8-bp DR sequences were clustered using CD-HIT (100% nucleotide identity and 100% coverage) 30 .The ggplot2 (ver.3.3.3)and ggseqlogo (ver.0.1) in R (ver.3.5.1)were used to analyze the GC content and nucleotide sequence conservation of target sequences, respectively.
In vitro experiments using a single-copy IS26 strain The construction and validation of the plasmid pUC57:IS26bla KPC-2 (5149 bp) were performed as previously described 12 .
Briefly, the pUC57:IS26-bla KPC-2 plasmid was constructed by assembling three fragments, including an IS26 gene, a bla KPC-2 gene, and a pUC57 vector, using an NEBuilder HiFi DNA assembly cloning kit.Then, pUC57:IS26-bla KPC-2 was transformed into E. coli MG1655 and coated onto LB agar plates containing kanamycin (50 μg/ml) and meropenem (2 μg/ml) for the selection of transformant.A single colony was picked from the transformation plate and inoculated with 200 μl of LB media in a 96-cell plate and incubated at 37°C to create a starter culture.Bacterial cells from the starter culture were then separately transferred to 200 μl of LB media containing meropenem or LB media without antibiotic (as control) per 12 h.The concentration of meropenem was gradiently increased by two times during each transfer, starting at 1 μg/ml, and doubled up to 256 μg/ml.In detail, the bacterial cells from the starter culture were inoculated to LB media with a carbapenem of 1 μg/ml.Then, the culture was further inoculated every 12 h, and the concentration of carbapenem was doubled each time until up to 256 μg/ml carbapenem.All the cultures were incubated at 37°C.Three biological replicates were performed.Then, the complete genome sequences for the whole culture treated with 256 μg/ml were obtained through the PromethION sequencing platform.
To detect the insertion site of the pUC57:IS26-bla KPC-2 plasmid, homogeneous search for IS26-bla KPC-2 against the complete genome we sequenced above was performed by BLASTN.

Figure 1 .
Figure 1.Construction and characterization of tIS26-bUs-ARGs.(A) The technical flow of the construction of the tIS26-bUs-ARGs data set.(B) Distribution of ARGs carried by tIS26-bUs.From inner to outer, the genes in the first ring indicate the ARGs carried by tIS26-bUs, the size of the circles in the second ring indicate the number of ARGs, the different colors in the third ring indicate the antibiotic classes for which resistance is conferred by the annotated ARGs, and the histograms in the fourth ring indicate the contribution rate of tIS26-bUs to ARGs.(C) The phylogenetic tree, constructed based on the genomes of the tIS26-bUs-ARGs-positive strains.From inner to outer, the first three rings indicate the taxonomy of each strain.The fourth ring shows the location of tIS26-bUs-ARGs (yellow, chromosome; green, plasmid; purple, both chromosome and plasmid).The fifth ring shows the number of tIS26-bUs-ARGs detected on each strain.The sixth ring shows the number of ARGs carried by each strain.The seventh ring shows the classes of the ARGs carried by each strain.(D) Venn diagram shows the relationship between the strain sets of plasmidlocated tIS26-bUs-ARGs and that of chromosome-located tIS26-bUs-ARGs.(E) The distribution of plasmid-and chromosome-located tIS26-bUs-ARGs of strains from different species.The red stars indicate the species of Enterobacteriaceae.ARGs, antimicrobial resistance genes; tIS26-bUs, translocatable IS26-bounded units; tIS26-bU-ARGs, translocatable IS26-bounded units with antimicrobial resistance genes.
Transfer capacity of the tIS26-bUs-ARGs.(A) The result of the genomic context environments of tIS26-bUs-ARGs in the same cluster.(B) The result of the distribution of tIS26-bUs-ARGs in the same cluster at different taxonomic levels.

Figure 3 .
Figure 3.The scatter plots of the proportion of tIS26-bUs-ARGs positive strains in sequenced strains over time.

Figure 4 .
Figure 4.The features of IS26 target site.(A) The GC content of IS26 target sites.(B) The consensus sequence of IS26 target sites.

Figure 5 .
Figure 5.The genomic position of the two random insertions of pUC57:IS26-bla KPC-2 in the chromosome of MG1655.