http://orcid.org/0000-0003-2005-9846
Figures
Abstract
Toxin-antitoxin systems (TAS) are commonly found on bacterial plasmids and are generally involved in plasmid maintenance. In addition to plasmid maintenance, several plasmid-mediated TAS are also involved in bacterial stress response and virulence. Even though the same TAS are present in a variety of plasmid types and bacterial species, differences in their sequences, expression and functions are not well defined. Here, we aimed to identify commonly occurring plasmid TAS in Escherichia coli and Klebsiella pneumoniae and compare the sequence, expression and plasmid stability function of their variants. 27 putative type II TAS were identified from 1063 plasmids of Klebsiella pneumoniae in GenBank. Among these, ccdAB and pemIK were found to be most common, also occurring in plasmids of E. coli. Comparisons of ccdAB variants, taken from E. coli and K. pneumoniae, revealed sequence differences, while pemIK variants from IncF and IncL/M plasmids were almost identical. Similarly, the expression and plasmid stability functions of ccdAB variants varied according to the host strain and species, whereas the expression and functions of pemIK variants were consistent among host strains. The specialised functions of some TAS may determine the host specificity and epidemiology of major antibiotic resistance plasmids.
Citation: Wu AY, Kamruzzaman M, Iredell JR (2020) Specialised functions of two common plasmid mediated toxin-antitoxin systems, ccdAB and pemIK, in Enterobacteriaceae. PLoS ONE 15(6): e0230652. https://doi.org/10.1371/journal.pone.0230652
Editor: Günther Koraimann, University of Graz, AUSTRIA
Received: March 4, 2020; Accepted: June 17, 2020; Published: June 30, 2020
Copyright: © 2020 Wu et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This work was funded by grant numbers G1084672 and G1145914, National Health and Medical Research Council (NHMRC), Australia to JRI and MK.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Toxin-antitoxin systems (TAS) were originally discovered on bacterial plasmids in the 1980s, [1, 2] but have since also been recognised on bacterial chromosomes [3]. TAS cassettes typically consist of two gene loci, governed by a common regulation mechanism [4], encoding a stable toxin that induces cell death or arrests growth, and a labile antitoxin that neutralises the toxin through binding to the toxin or other means. While the toxin is always a protein, the antitoxin can be protein or RNA based, and thus TAS can be categorised into six different types (Types I-VI), based on the nature and mechanism of action of the antitoxin [5, 6]. The Type II system, in which both toxin and antitoxin are proteins, is the typical model of a TAS and the best studied, and is probably the most common in bacteria [7].
It has long been known that plasmid-mediated TAS play a role in plasmid maintenance through postsegregational killing and/or inhibition of growth of plasmid free cells [2, 8, 9], while chromosomal TAS have been found to be involved in a range of other roles, such as in the bacterial stress response, antibiotic tolerance, persister cell formation, biofilm formation, bacterial virulence and intestinal colonisation [10–14]. However, several recent studies identified the involvement of plasmid-mediated TAS in a range of bacterial physiology beyond plasmid maintenance. We have recently showed that the plasmid-mediated parDE TA system is involved in antibiotic and heat tolerance in Escherichia coli [15]. Some other plasmid-mediated TA systems such as ccdAB contributes antibiotic persistence in E. coli [16], while mpvAT is involved in virulence by maintaining virulence plasmids in Shigella and Salmonella species [17, 18]. Within Enterobacteriaceae, TAS are common among conjugative plasmids including antibiotic resistance (AbR) plasmids, and often associated with certain plasmid incompatibility (Inc) types. For example, the type II TAS pemIK and vagCD are usually found on IncF and IncL/M plasmids, and IncF and IncHI2 plasmids respectively [19, 20].
The distribution of TAS in the plasmids found in Escherichia coli is well described [21–25], but there has been as yet no comprehensive study of the TAS in all known plasmids residing in Klebsiella pneumoniae species. Existing studies of a collection of clinically relevant K. pneumoniae genomes (encompassing both plasmids and chromosomes) has provided insights into the distribution of TAS in the chromosomes of this species and the association of some TAS with certain plasmid replicon and antibiotic resistance regions, suggesting possible strain specific specialisation of TAS, with some being localised to one species while others are found across species [26, 27]. Previous studies have also identified two distinct groups of vagCD TAS on plasmids located in K. pneumoniae [28], with each of the toxins and antitoxins shown to be functional. vagCD is a member of the large TA family vapBC, with the toxin VagD being a PilT N-terminal (PIN) domain containing endoribonuclease that inhibits translation [29, 30].
TAS with identical names infer similar roles, regardless of the initial source and sequence variations, but functional differences in TAS from different genetic contexts have previously been noted [31, 32]. For example, IncF plasmid-mediated ccdAB TAS (ccdF) from pathogenic E. coli O157:H7 is involved in post-segregational killing but its chromosomal counterpart (ccdO157) is unable to mediate PSK [32]. However, common TAS functions (e.g. plasmid maintenance or antibiotic tolerance) for plasmid-mediated TA variants were not investigated.
In this study, we analysed the distribution of type II TAS in 1063 fully sequenced plasmids found in K. pneumoniae retrieved from GenBank. Variation in DNA sequence, expression and function of two common TAS found in both E. coli and K. pneumoniae plasmids, ccdAB and pemIK, were examined.
ccdAB is a well-studied, primarily plasmid-associated type II TAS, although copies have also been found on bacterial chromosomes where it appears to be involved in the bacterial stress response [33, 34]. The toxin, CcdB, binds to and disrupts the action of DNA gyrase, causing double strand DNA breaks and the induction of the bacterial SOS response [35, 36]. pemIK, another well studied type II TAS that is related to the mazEF system, is an mRNA endoribonuclease that inhibits protein synthesis [37]. Here, we provide vital information about the similarity, specificity and functions of these two TAS, with broad implications for their role in the spread of antibiotic resistance.
Materials and methods
Identification of TAS on K. pneumoniae plasmids
The names and complete sequences of all plasmids from K. pneumoniae species available in GenBank at the time of search (August 2019) were retrieved (https://www.ncbi.nlm.nih.gov/genome/?term=klebsiella+pneumoniae), and the sequences of all plasmids >30 kb examined with TA Finder (http://202.120.12.133/TAfinder/TAfinder.php) [38] using the default parameters to identify potential type II TAS. Plasmids <30 kb in size were excluded as smaller plasmids generally do not carry TA systems. The plasmid incompatibility (Inc) type of each plasmid was defined using PlasmidFinder (https://cge.cbs.dtu.dk/services/PlasmidFinder/) [39].
Alignment of TAS sequences
Representative examples of the two most common TAS, ccdAB and pemIK, were chosen, and the nucleotide and amino acid sequences of the toxin and antitoxin coding regions retrieved from GenBank (https://www.ncbi.nlm.nih.gov/genbank/) using coordinates obtained from TA Finder. To verify that these sequences were in fact representative of their specific variant, each of the sequences were used as queries for nucleotide Basic Local Alignment Search Tool (BLAST) searches (https://blast.ncbi.nlm.nih.gov/Blast.cgi) [40]. The resulting matches were compiled for each TAS, then aligned in MEGA7 (http://www.megasoftware.net/mega7/) [41] using the ClustalW algorithm, followed by construction of phylogenetic trees using the Maximum Likelihood method.
Variants were compared with one another by nucleotide alignment as described above. The amino acid sequences were then used to predict the secondary structures of the proteins using PsiPred (http://bioinf.cs.ucl.ac.uk/psipred/) [42].
TAS promoters were predicted using BPROM (http://www.softberry.com/berry.phtml?topic=bprom&group=programs&subgroup=gfindb) (Softberry) with default parameters, with the input being the 500 bp region upstream of the ATG start codon of the antitoxin gene. The putative promoter sequences were then aligned as described above.
Plasmids, bacteria, primers and culture conditions
Tables 1, 2 and 3 list plasmids, bacterial strains and primers respectively. Bacteria were grown in Luria-Bertani (LB) broth (BD Biosciences, NJ, USA), with kanamycin (50 μg/mL) or chloramphenicol (20 μg/mL) (Sigma-Aldrich, MO, USA) added as indicated. Insertion of TAS into vectors was carried out using standard restriction digestion and ligation cloning protocols, and chemical transformation and electroporation into host strains was also performed using standard protocols. The plasmid construct details can be found in Table 1. Each solution used was rendered sterile either through autoclaving or filter sterilising at 0.22 μm.
Measurement of relative promoter strength
Relative strengths of putative TAS promoters were determined from TA promoter-gfp constructs using methods described previously [47]. Briefly, the predicted TAS promoters were cloned upstream of a promoterless gfp in the expression vector pANT3, and four strains each of E. coli and K. pneumoniae were transformed with these constructs. Overnight cultures were inoculated from single colonies into LB broth with kanamycin, and grown with shaking at 37°C. The cultures were then diluted 200 x in LB broth and grown under the same conditions for a further 3–4 h. Cells were harvested by centrifugation and the pellets resuspended in 0.85% sodium chloride (saline). The concentrations of each sample were then standardised to OD580 of 1.0 (~ 8.0 x 108 cfu/mL), as determined using the DensiCHEKTM Plus nephelometer (bioMérieux, France). Fluorescence was analysed using a Victor3 plate reader (Perkin Elmer, MA, USA), with an excitation wavelength of 485 nm and an emission wavelength of 535 nm. Experiments were performed in triplicate, and readings were averaged and corrected for background fluorescence by subtracting the pANT3 (no promoter) reading from each sample. The positive control pANT5, which constitutively expresses GFP from the ptac promoter, is a derivative of pANT3 [43]. Both plasmids carry the pRSF1010 origin of replication, which is a low copy broad-host range IncQ plasmid [48].
Plasmid stability assays
To assess plasmid stability, the whole TAS (including the putative promoter and ribosome binding site) was cloned into a low and a high copy plasmid (pACYC184 and pBSCK+ respectively) and two strains each of E. coli and K. pneumoniae were transformed with these constructs. Plasmid stability was assessed as described previously [15]. Briefly, a single colony of E. coli or K. pneumoniae bacteria carrying the relevant plasmid was grown in LB broth at 37°C with shaking at 225 rpm for 72 or 96 hours, without antibiotic selection. Bacterial cultures were transferred into fresh LB medium at 1:1000 dilution at 16, 24, 40, 48, 64 and 72 hrs for high copy plasmids. Additional dilutions were performed at 88 and 96 hrs for low copy plasmids, as these are not lost as quickly as the high copy plasmids. Samples were taken before every transfer, diluted in saline and plated on to LB agar without antibiotic and incubated at 37°C for 18 h. From each plate, 120 colonies were replica plated onto LB agar plates with and without the indicator antibiotics to estimate plasmid retention.
Results and discussion
Distribution of type II TAS in the plasmids found in K. pneumoniae strains
Twenty-seven different putative type II TAS were identified (Table 4), with ccdAB, pemIK and vagCD most common among them (S1 Table). We compared the distribution of the TAS identified here in K. pneumoniae plasmids with previously reported TAS in E. coli plasmids [21]. E. coli and K. pneumoniae are closely related members of the Enterobacteriaceae family, sharing a large number of mobile antibiotic resistance genes, mainly via plasmids. However, only three TAS (ccdAB, pemIK, and vagCD) were common in both, with ccdAB and pemIK most commonly shared. Many, such as MNT-HEPN-like, GNAT-RHH-like, Bro-Abr-like and Bro-ArsR-like TAS appear to be most important in Klebsiella plasmids and have not previously been reported in E. coli plasmids [21, 23].
Plasmids found in K. pneumoniae species belonged to 11 different Inc types (Table 4), although around 10% (114/1063) were not assigned an Inc type in PlasmidFinder. More than half of the plasmids were from the IncF replicon group (Fig 1), which is the most common plasmid replicon type in Enterobacteriaceae [49]. Replication gene variation can be used to further subdivide IncF plasmids (e.g. IncFIA, IncFIB, IncFIC, IncFII etc) and IncFII plasmids are usually further subdivided with a subscript to indicate the typical host species, e.g. IncFIIK and IncFIIS for the IncFII plasmids found in Klebsiella and Salmonella species, respectively. For simplicity, all IncF plasmid subtypes found in K. pneumoniae strains were grouped as IncF in this study. A number of K. pneumoniae plasmid types (L, M, X, A, C, N and H) (Table 4, Fig 1) are also commonly found in E. coli and other Enterobacteriaceae, [50] but some that are quite commonly reported in K. pneumoniae (e.g. IncR, representing 76/1083 plasmids) are rarely reported in E. coli.
Plasmid incompatibility types are referred to as Inc and ND refers to plasmids that could not be typed by Plasmid Finder. The TAS commonly found in each plasmid type are shown in italics.
Other associations are noteworthy. For example, ccdAB is somewhat specific to IncF plasmids in K. pneumoniae, while pemIK is common to both IncL/M and IncF plasmids (Table 4, Fig 1, S1 Table), in line with previous observations [19]. vagCD is another TAS predominantly found on the blaCTX-M-carrying IncF plasmids in E. coli (also as previously noted) [21], and in K. pneumoniae (Table 4, Fig 1). However, vagCD was also common in IncH (37/58) and IncR (35 of 76) plasmids in K. pneumoniae but not in E. coli. Most IncX plasmids lacked an easily identifiable TAS, save for a few carrying hicBA, also as previously noted [51].
ccdAB sequences differ between species, while pemIK is highly conserved
ccdAB and pemIK were the two most predominant TAS modules across both E. coli and K. pneumoniae plasmids (Table 4, [21]), and were thus chosen for further study.
Representative sequences of ccdAB were chosen from IncF plasmids derived from two species (E. coli and K. pneumoniae; henceforth termed “ccdABEC” and “ccdABKP” respectively), while variants of pemIK were chosen from two plasmid incompatibility types, IncL/M and IncF (henceforth termed “pemIKLM” and “pemIKF” respectively). These sequences were confirmed as representative of their specific variant by phylogenetic analysis (Fig 2). ccdAB shows two clear groups, with one dominated by E. coli and one by K. pneumoniae, representing ccdABEC and ccdABKP respectively (Fig 2A). Each variant was also found, in rare instances, in other species, however given that these are self-transmissible conjugative plasmids, this is to be expected. Similarly, pemIK also branched into two groups, with one being dominated by IncF plasmids, and the other by IncL/M plasmids (Fig 2B), irrespective to species. These results confirm that each variant chosen is representative of their specific group.
Phylogenetic trees of (A) ccdAB and (B) pemIK. Representative sequences were used as queries in BLAST searches. Randomly chosen examples of the top 50 resulting hits were then aligned in MEGA7 using the ClustalW algorithm and phylogenetic trees constructed using the Maximum Likelihood Method.
Alignment of ccdABEC and ccdABKP revealed 85% identity in the nucleotide coding sequence (S1A Fig), translating to 75% and 93% amino acid sequence identities in the antitoxin and toxin respectively (Fig 3A). Predicted secondary structure analysis indicated that although the antitoxin CcdA had the same basic structure in both cases, the toxin CcdB derived from K. pneumoniae had an additional small α-helix near the C-terminal end (S2 Fig). The putative promoters were less similar, with only a 33% identity throughout the promoter region (-10, -35 and spacers) (Fig 3C). Thus, it is expected that the expression of these TA modules in different species would vary. It has previously been reported that differences in the promoter regions of chromosomal and plasmid mediated ccdAB variants was associated with differences in expression and function [32].
These include the amino acid sequence alignments of (A) the antitoxin CcdA and the toxin CcdB, from E. coli and K. pneumoniae; and (B) the antitoxin PemI and the toxin PemK, from IncL/M and IncF plasmids; and (C) the nucleotide alignments of the putative promoter regions of the selected TAS. The putative -35 and -10 elements are boxed. All sequences were aligned in MEGA7 using the ClustalW algorithm. Non-identical residues/bases are highlighted in black.
By contrast, pemIK sequences differed with the plasmid Inc types, rather than with the host bacterial species. Alignment of pemIKLM pemIKF revealed 92% identity in the nucleotide coding sequence (S1B Fig), translating to 95% amino acid sequence identities in both proteins (Fig 3B). The PemI antitoxins and PemK toxins also had identical secondary structures (S3 Fig). Similarly, the putative promoters had 94% nucleotide identity (Fig 3C) with only one SNP in the -10 sequences, indicating that pemIK TAS is quite conserved within different plasmid types.
Variation in expression of ccdAB in different host strains is greater than expression of pemIK
GFP expression from both pemIK promoter variants were relatively strong in all strains/species tested, with expression higher than from either of the ccdAB promoters tested, as well as the workhorse tac promoter [52] (Fig 4), in line with the apparently minor variation in pemIK promoters. pemIK is the sole TAS evident in published IncL/M plasmids [19], and these data suggest that pemIK alone is sufficient to provide stability to the plasmids carrying them, regardless of the host bacterial species.
Relative expression of GFP from promoters of (A) ccdAB and (B) pemIK. The putative promoter of each TAS was inserted upstream of promoterless gfp in pANT3 and transformed into four E. coli and four K. pneumoniae strains. pANT5, with a constitutive tac promoter (ptac)-gfp construct, served as a positive control, and values are corrected for background noise. Data shown are the means of three replicates, with the error bars representing one standard deviation from the mean. Please note the ten-fold differences in y-axis scales.
In contrast, expression of GFP from ccdAB promoters is generally lower than that of pemIK and showed host species specificity. GFP expression from the E. coli specific ccdAB promoter was higher in E. coli than in K. pneumoniae strains. The expression of GFP from the K. pneumoniae specific ccdAB promoter also varied between strains in each species but was relatively lower than from all other promoters tested here. The association between differences in expression of ccdAB from different genetic contexts with differences in function in E. coli has previously been noted [32].
ccdAB function is specific to host strain, whereas pemIK is not
To assess whether the apparent specialisations within ccdAB (by host species) and pemIK (by plasmid type) are also reflected functionally, we assessed plasmid stability in two host strains of each species, in both low and high copy number plasmids. The low copy plasmid data are presented (Fig 5) as the natural plasmids carrying these TAS are generally low copy number. Consistent data for high copy plasmids can be found in the supplementary material (S4 Fig).
Each TAS was cloned into a pACYC184 backbone, and the percentage of cells retaining the plasmid calculated at each time point. Measurements were done in two species, E. coli (A: Ec BW25113 and B: Ec WH59) and K. pneumoniae (C: Kp ATCC13883 and D: Kp WH49). Data shown are the means of three replicates, with the error bars indicating one standard deviation from the mean.
The plasmid stability conferred by ccdAB variants varies between host species (Fig 5) and is generally consistent with the GFP expression data, with ccdABEC stabilising plasmids in both E. coli strains but not in K. pneumoniae strains. ccdABKP, on the other hand, appears to be less specialised, conferring some plasmid stability in all strains, regardless of species. Given the relatively low level of expression from ccdAB promoters and discrepancies in the conferred plasmid stability across species and strain, it is perhaps unsurprising that it is typically one of several TAS in IncF plasmids [21]. By contrast, the more reliably and vigorously expressed pemIK variants conferred significant plasmid stability in all strains tested (Fig 5).
Overall, pemIK appears less specialised than ccdAB. The sequences and expression of pemIK variants differed relatively less, and plasmid stability functions of both variants were more consistent. In contrast, ccdAB appears to be more variable in sequence and function, different variants conferring plasmid stability in different strains.
ccdAB is one of the most well studied TAS and has been reported on both plasmids and chromosomes of different bacterial species [33, 34]. Their sequences and functions also varied based on their genetic context or source of origin (plasmids and bacterial species), and functional differences between plasmid-mediated and chromosomal ccdAB have been demonstrated previously [32]. The finding that the CcdB toxin found in the Salmonella virulence plasmid pSLT, which has a single amino acid substitution (R99W), is non-toxic and unable to provide plasmid maintenance [53], suggests that even minor variations in TAS protein sequences may lead to changes in their functions. It has also been demonstrated that the antitoxin CcdA originating from one source does not necessarily protect bacteria from toxicity mediated by the toxin CcdB from a different origin, even though both toxins followed the same mechanism of action. For example, CcdA from Vibrio fischeri superintegron could not protect from the toxicity of plasmid mediated CcdBF [54].
The inability of ccdAB to confer significant plasmid stability in certain host strains raises the possibility that it is performing another role in these strains. It would be of interest to investigate whether these plasmid-mediated ccdAB variants are involved in other known functions of TAS within these strains, such as persister formation, antibiotic and heat tolerance, and other bacterial stress responses [10, 12, 15, 33, 55, 56]. Within E. coli, both an IncF plasmid mediated ccdAB and chromosomally located variant have previously been shown to be able to confer some level of protection during antibiotic treatment [33], and the toxin CcdB from IncF plasmid has also been linked to increases in persister cell formation [16].
Variation in ccdAB has previously been noted to influence the spread of certain Salmonella serovars [53], and we believe that the variation in expression and plasmid stability function of ccdAB and pemIK may contribute to and influence the epidemiology of the conjugative antibiotic resistance plasmids which carry them. It is hoped that an increased understanding of these effects may be able to aid in the surveillance of these plasmids and the spread of the resistance genes carried by them.
Supporting information
S1 Table. Distribution of type II TA systems in the plasmids of K. pneumonia.
https://doi.org/10.1371/journal.pone.0230652.s001
(XLSX)
S1 Fig. The alignment and comparison of the nucleotide sequences of ccdAB and pemIK.
Nucleotide sequence alignments of the coding region of (A) ccdAB from plasmids found in E. coli and K. pneumoniae, and (B) pemIK from IncL/M and IncF plasmids. Sequences were aligned in MEGA7 using the ClustalW algorithm. Non-identical residues are highlighted in black.
https://doi.org/10.1371/journal.pone.0230652.s002
(TIF)
S2 Fig. Comparison of the predicted secondary structures of ccdAB.
The predicted secondary structure of the ccdAB encoded toxins and antitoxins from plasmids found in E. coli and K. pneumoniae. Arrows represent β-strands, cylinder shapes represent α-helices and lines represent random coils. Pred: predicted secondary structure; AA: amino acids; numbers below each structure represent the amino acid positions within the proteins.
https://doi.org/10.1371/journal.pone.0230652.s003
(TIF)
S3 Fig. Comparison of the predicted secondary structures of pemIK.
The predicted secondary structure of the pemIK encoded toxins and antitoxins from IncL/M and IncF type plasmids. Arrows represent β-strands, cylinder shapes represent α-helices and lines represent random coils. Pred: predicted secondary structure; AA: amino acids; numbers below each structure represent the amino acid positions within the proteins.
https://doi.org/10.1371/journal.pone.0230652.s004
(TIF)
S4 Fig. Plasmid stability effects of ccdAB and pemIK in a high copy plasmid.
Stability of a high copy number plasmid with and without TAS over 72 h. Each TAS was cloned into a pBCSK+ backbone, and the percentage of cells retaining the plasmid calculated at each time point. Measurements were done in two species, E. coli (A: Ec BW25113 and B: Ec WH59) and K. pneumoniae (C: Kp ATCC13883 and D: Kp WH49). Data shown are the means of three replicates, with the error bars indicating one standard deviation from the mean.
https://doi.org/10.1371/journal.pone.0230652.s005
(TIF)
References
- 1. Jaffe A, Ogura T, Hiraga S. Effects of the ccd function of the F plasmid on bacterial growth. J Bacteriol. 1985;163(3):841–9. Epub 1985/09/01. pmid:3897195.
- 2. Gerdes K, Rasmussen PB, Molin S. Unique type of plasmid maintenance function: postsegregational killing of plasmid-free cells. Proc Natl Acad Sci U S A. 1986;83(10):3116–20. Epub 1986/05/01. pmid:3517851.
- 3. Gotfredsen M, Gerdes K. The Escherichia coli relBE genes belong to a new toxin-antitoxin gene family. Mol Microbiol. 1998;29(4):1065–76. pmid:9767574
- 4. Engelberg-Kulka H, Glaser G. Addiction modules and programmed cell death and antideath in bacterial cultures. Annu Rev Microbiol. 1999;53:43–70. Epub 1999/11/05. pmid:10547685.
- 5. Brantl S, Jahn N. sRNAs in bacterial type I and type III toxin-antitoxin systems. FEMS Microbiol Rev. 2015;39(3):413–27. pmid:25808661.
- 6. Aakre CD, Phung TN, Huang D, Laub MT. A bacterial toxin inhibits DNA replication elongation through a direct interaction with the beta sliding clamp. Mol Cell. 2013;52(5):617–28. pmid:24239291.
- 7. Chan WT, Espinosa M, Yeo CC. Keeping the wolves at bay: Antitoxins of prokaryotic type II toxin-antitoxin systems. Front Mol Biosci. 2016;3:9. pmid:27047942.
- 8. Hayes F. Toxins-antitoxins: plasmid maintenance, programmed cell death, and cell cycle arrest. Science. 2003;301(5639):1496–9. Epub 2003/09/13. pmid:12970556.
- 9. Song S, Wood TK. Post-segregational killing and phage inhibition are not mediated by cell death through Toxin/Antitoxin systems. Front Microbiol. 2018;9:814. Epub 2018/06/21. pmid:29922242.
- 10. Wang X, Wood TK. Toxin-antitoxin systems influence biofilm and persister cell formation and the general stress response. Appl Environ Microbiol. 2011;77(16):5577–83. pmid:21685157.
- 11. Kasari V, Kurg K, Margus T, Tenson T, Kaldalu N. The Escherichia coli mqsR and ygiT genes encode a new toxin-antitoxin pair. J Bacteriol. 2010;192(11):2908–19. Epub 2010/03/18. pmid:20233923.
- 12. Gerdes K, Christensen SK, Lobner-Olesen A. Prokaryotic toxin-antitoxin stress response loci. Nat Rev Microbiol. 2005;3(5):371–82. pmid:15864262.
- 13. Gerdes K, Maisonneuve E. Bacterial persistence and toxin-antitoxin loci. Annu Rev Microbiol. 2012;66:103–23. Epub 2012/09/22. pmid:22994490.
- 14. Wang Y, Wang H, Hay AJ, Zhong Z, Zhu J, Kan B. Functional RelBE-Family toxin-antitoxin pairs affect biofilm maturation and intestine colonization in Vibrio cholerae. PLoS One. 2015 2015; 10(8):[e0135696 p.].
- 15. Kamruzzaman M, Iredell J. A ParDE-family toxin antitoxin system in major resistance plasmids of Enterobacteriaceae confers antibiotic and heat tolerance. Sci Rep. 2019;9(1):9872. Epub 2019/07/10. pmid:31285520.
- 16. Tripathi A, Dewan PC, Barua B, Varadarajan R. Additional role for the ccd operon of F-plasmid as a transmissible persistence factor. Proc Natl Acad Sci U S A. 2012;109(31):12497–502. Epub 2012/07/18. pmid:22802647.
- 17. Pilla G, Tang CM. Going around in circles: virulence plasmids in enteric pathogens. Nat Rev Microbiol. 2018;16(8):484–95. Epub 2018/06/02. pmid:29855597.
- 18. McVicker G, Hollingshead S, Pilla G, Tang CM. Maintenance of the virulence plasmid in Shigella flexneri is influenced by Lon and two functional partitioning systems. Mol Microbiol. 2019;111(5):1355–66. Epub 2019/02/16. pmid:30767313.
- 19. Kamruzzaman M, Shoma S, Thomas CM, Partridge SR, Iredell JR. Plasmid interference for curing antibiotic resistance plasmids in vivo. PLoS One. 2017;12(2):e0172913. pmid:28245276.
- 20. Mnif B, Vimont S, Boyd A, Bourit E, Picard B, Branger C, et al. Molecular characterization of addiction systems of plasmids encoding extended-spectrum β-lactamases in Escherichia coli. J Antimicrob Chemother. 2010;65(8):1599–603. pmid:20507859.
- 21. Doumith M, Dhanji H, Ellington MJ, Hawkey P, Woodford N. Characterization of plasmids encoding extended-spectrum β-lactamases and their addiction systems circulating among Escherichia coli clinical isolates in the UK. J Antimicrob Chemother. 2012;67(4):878–85. pmid:22210753.
- 22. Tsilibaris V, Maenhaut-Michel G, Mine N, Van Melderen L. What is the benefit to Escherichia coli of having multiple toxin-antitoxin systems in its genome? J Bacteriol. 2007;189(17):6101–8. pmid:17513477
- 23. Yamaguchi Y, Park JH, Inouye M. Toxin-antitoxin systems in bacteria and archaea. Annu Rev Genet. 2011;45:61–79. pmid:22060041.
- 24. Fiedoruk K, Daniluk T, Swiecicka I, Sciepuk M, Leszczynska K. Type II toxin-antitoxin systems are unevenly distributed among Escherichia coli phylogroups. Microbiology. 2015;161(Pt 1):158–67. Epub 2014/11/08. pmid:25378561.
- 25. Ramisetty BC, Santhosh RS. Horizontal gene transfer of chromosomal Type II toxin-antitoxin systems of Escherichia coli. FEMS Microbiol Lett. 2016;363(3). Epub 2015/12/17. pmid:26667220.
- 26. Horesh G, Fino C, Harms A, Dorman MJ, Parts L, Gerdes K, et al. Type II and type IV toxin-antitoxin systems show different evolutionary patterns in the global Klebsiella pneumoniae population. Nucleic Acids Res. 2020. Epub 2020/04/02. pmid:32232417.
- 27. Wei YQ, Bi DX, Wei DQ, Ou HY. Prediction of Type II Toxin-Antitoxin Loci in Klebsiella pneumoniae Genome Sequences. Interdiscip Sci. 2016;8(2):143–9. Epub 2015/12/15. pmid:26662948.
- 28. Duprilot M, Decre D, Genel N, Drieux L, Sougakoff W, Arlet G. Diversity and functionality of plasmid-borne vagCD toxin-antitoxin systems of Klebsiella pneumoniae. J Antimicrob Chemother. 2017;72(5):1320–6. Epub 2017/01/26. pmid:28119479.
- 29. Arcus VL, McKenzie JL, Robson J, Cook GM. The PIN-domain ribonucleases and the prokaryotic VapBC toxin-antitoxin array. Protein Eng Des Sel. 2011;24(1–2):33–40. Epub 2010/11/03. pmid:21036780.
- 30. Maisonneuve E, Shakespeare LJ, Jorgensen MG, Gerdes K. Bacterial persistence by RNA endonucleases. Proc Natl Acad Sci U S A. 2011;108(32):13206–11. Epub 2011/07/27. pmid:21788497.
- 31. Harms A, Brodersen DE, Mitarai N, Gerdes K. Toxins, Targets, and Triggers: An overview of toxin-antitoxin biology. Mol Cell. 2018;70(5):768–84. pmid:29398446
- 32. Wilbaux M, Mine N, Guerout AM, Mazel D, Van Melderen L. Functional interactions between coexisting toxin-antitoxin systems of the ccd family in Escherichia coli O157:H7. J Bacteriol. 2007;189(7):2712–9. Epub 2007/01/30. pmid:17259320.
- 33. Gupta K, Tripathi A, Sahu A, Varadarajan R. Contribution of the chromosomal ccdAB operon to bacterial drug tolerance. J Bacteriol. 2017;199(19). Epub 2017/07/05. pmid:28674066.
- 34. Pandey DP, Gerdes K. Toxin-antitoxin loci are highly abundant in free-living but lost from host-associated prokaryotes. Nucleic Acids Res. 2005;33(3):966–76. pmid:15718296.
- 35. Couturier M, Bahassi el M, Van Melderen L. Bacterial death by DNA gyrase poisoning. Trends Microbiol. 1998;6(7):269–75. Epub 1998/08/26. pmid:9717215.
- 36. Bernard P, Couturier M. Cell killing by the F plasmid CcdB protein involves poisoning of DNA-topoisomerase II complexes. J Mol Biol. 1992;226(3):735–45. Epub 1992/08/05. pmid:1324324.
- 37. Zhang J, Zhang Y, Zhu L, Suzuki M, Inouye M. Interference of mRNA function by sequence-specific endoribonuclease PemK. J Biol Chem. 2004;279(20):20678–84. Epub 2004/03/17. pmid:15024022.
- 38. Shao Y, Harrison EM, Bi D, Tai C, He X, Ou HY, et al. TADB: a web-based resource for Type 2 toxin-antitoxin loci in bacteria and archaea. Nucleic Acids Res. 2011;39(Database issue):D606–11. Epub 2010/10/12. pmid:20929871.
- 39. Carattoli A, Zankari E, Garcia-Fernandez A, Voldby Larsen M, Lund O, Villa L, et al. In silico detection and typing of plasmids using PlasmidFinder and plasmid multilocus sequence typing. Antimicrob Agents Chemother. 2014;58(7):3895–903. Epub 2014/04/30. pmid:24777092.
- 40. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol. 1990;215(3):403–10. Epub 1990/10/05. pmid:2231712.
- 41. Kumar S, Stecher G, Tamura K. MEGA7: Molecular Evolutionary Genetics Analysis version 7.0 for bigger datasets. Mol Biol Evol. 2016;33(7):1870–4. Epub 2016/03/24. pmid:27004904.
- 42. Jones DT. Protein secondary structure prediction based on position-specific scoring matrices. J Mol Biol. 1999;292(2):195–202. Epub 1999/09/24. pmid:10493868.
- 43. Lee AK, Falkow S. Constitutive and inducible green fluorescent protein expression in Bartonella henselae. Infect Immun. 1998;66(8):3964–7. Epub 1998/07/23. pmid:9673287.
- 44. Partridge SR, Ginn AN, Paulsen IT, Iredell JR. pEl1573 Carrying blaIMP-4, from Sydney, Australia, is closely related to other IncL/M plasmids. Antimicrob Agents Chemother. 2012;56(11):6029–32. Epub 2012/08/29. pmid:22926566.
- 45. Grenier F, Matteau D, Baby V, Rodrigue S. Complete genome sequence of Escherichia coli BW25113. Genome Announc. 2014;2(5). Epub 2014/10/18. pmid:25323716.
- 46. Agyekum A, Fajardo-Lubian A, Ai X, Ginn AN, Zong Z, Guo X, et al. Predictability of phenotype in relation to common β-lactam resistance mechanisms in Escherichia coli and Klebsiella pneumoniae. J Clin Microbiol. 2016;54(5):1243–50. Epub 2016/02/26. pmid:26912748.
- 47. Kamruzzaman M, Patterson JD, Shoma S, Ginn AN, Partridge SR, Iredell JR. Relative strengths of promoters provided by common mobile genetic elements associated with resistance gene expression in gram-negative bacteria. Antimicrob Agents Chemother. 2015;59(8):5088–91. pmid:26055385.
- 48. Meyer R. Replication and conjugative mobilization of broad host-range IncQ plasmids. Plasmid. 2009;62(2):57–70. pmid:19465049.
- 49. Carattoli A. Plasmids in Gram negatives: molecular typing of resistance plasmids. Int J Med Microbiol. 2011;301(8):654–8. Epub 2011/10/14. pmid:21992746.
- 50. Carattoli A. Resistance plasmid families in Enterobacteriaceae. Antimicrob Agents Chemother. 2009;53(6):2227–38. Epub 2009/03/25. pmid:19307361.
- 51. Bustamante P, Iredell JR. Carriage of type II toxin-antitoxin systems by the growing group of IncX plasmids. Plasmid. 2017;91:19–27. Epub 2017/03/08. pmid:28267580.
- 52. de Boer HA, Comstock LJ, Vasser M. The tac promoter: a functional hybrid derived from the trp and lac promoters. Proc Natl Acad Sci U S A. 1983;80(1):21–5. Epub 1983/01/01. pmid:6337371.
- 53. Di Cesare A, Losasso C, Barco L, Eckert EM, Conficoni D, Sarasini G, et al. Diverse distribution of Toxin-Antitoxin II systems in Salmonella enterica serovars. Sci Rep. 2016;6:28759. pmid:27357537.
- 54. Guerout AM, Iqbal N, Mine N, Ducos-Galand M, Van Melderen L, Mazel D. Characterization of the phd-doc and ccd toxin-antitoxin cassettes from Vibrio superintegrons. J Bacteriol. 2013;195(10):2270–83. Epub 2013/03/12. pmid:23475970.
- 55. Kedzierska B, Hayes F. Emerging roles of toxin-antitoxin modules in bacterial pathogenesis. Molecules. 2016;21(6). Epub 2016/06/21. pmid:27322231.
- 56. Lobato-Marquez D, Diaz-Orejas R, Garcia-Del Portillo F. Toxin-antitoxins and bacterial virulence. FEMS Microbiol Rev. 2016;40(5):592–609. Epub 2016/08/01. pmid:27476076.