https://orcid.org/0000-0002-0158-850X
Figures
Abstract
Escherichia coli exhibit extensive genetic diversity at the genome level, particularly within their accessory genome. The tRNA integrated genomic islands (GIs), a part of the E. coli accessory genome, play an important role in pathogenicity. However, studies examining the evolution of GIs have been challenging due to their large size, considerable gene content variation and fragmented assembly in draft genomes. Here we examined the evolution of the GI integrated at pheV-tRNA (GI-pheV), with a primary focus on uropathogenic E. coli (UPEC) and the globally disseminated multidrug resistant ST131 clone. We show the gene content of GI-pheV is highly diverse and arranged in a modular configuration, with the P4 integrase encoding gene intP4 the only conserved gene. Despite this diversity, the GI-pheV gene content displayed conserved features among strains from the same pathotype. In ST131, GI-pheV corresponding to the reference strain EC958 (EC958_GI-pheV) was found in ~90% of strains. Phylogenetic analyses suggested that GI-pheV in ST131 has evolved together with the core genome, with the loss/gain of specific modules (or the entire GI) linked to strain specific events. Overall, we show GI-pheV exhibits a dynamic evolutionary pathway, in which modules and genes have evolved through multiple events including insertions, deletions and recombination.
Author summary
Genomic islands (GIs) are key contributors to E. coli pathogenesis. However, studies on the evolution of GIs are limited, mostly due to their large size, variation, and fragmented assembly in draft genomes. Here we investigated the evolution of the GI-pheV genomic island. We showed that the presence of GI-pheV is primarily associated with pathogenic E. coli, with the greatest diversity in gene content observed among strains from different pathotypes. The GI-pheV in UPEC contains uropathogenicity-associated virulence and fitness encoding modules, which have either co-evolved with the island or evolved independently, possibly due to different selective pressures. Focused investigation of GI-pheV in ST131, the most globally dominant antibiotic resistant UPEC clone, revealed loss or gain of genes and modules via recombination during its evolution.
Citation: Nhu NTK, Forde BM, Ben Zakour NL, Phan M-D, Roberts LW, Beatson SA, et al. (2024) Evolution of the pheV-tRNA integrated genomic island in Escherichia coli. PLoS Genet 20(10): e1011459. https://doi.org/10.1371/journal.pgen.1011459
Editor: Diarmaid Hughes, Uppsala University, SWEDEN
Received: May 12, 2024; Accepted: October 14, 2024; Published: October 24, 2024
Copyright: © 2024 Nhu 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: PacBio read data for the complete ST131 genomes described in this study has been deposited at NCBI SRA (SRR10761087 (strain S37EC), SRR10761088 (strain S79EC), SRR10761089 (strain HVM2044)) under Bioproject PRJNA596846. The underlying numerical data for all graphs and summary statistics is provided in S1, S2 and S3 Tables.
Funding: This work was supported by grants from the Australian National Health and Medical Research Council (1181958 and 2001431 to MAS, M-DP, NTKN) and the Australian Infectious Diseases Research Centre (to MAS, NTKN). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Escherichia coli is an extremely diverse species. Some strains can colonize the gastrointestinal tract as harmless commensals, while others possess distinct sets of virulence genes that enable them to cause disease [1,2]. E. coli is classified based on virulence and site of infection, ranging from commensal strains to intestinal pathogenic E. coli (InPEC) and extra-intestinal pathogenic E. coli (ExPEC). InPEC and ExPEC can be further grouped into a range of pathotypes: InPEC—enteropathogenic E. coli (EPEC), enterohemorrhagic E. coli (EHEC), enterotoxigenic E. coli (ETEC), enteroaggregative E. coli (EAEC), enteroinvasive E. coli (EIEC), adherent-invasive E. coli (AIEC) and enteroaggregative and hemorrhagic E. coli (EAHEC); ExPEC—uropathogenic E. coli (UPEC), neonatal meningitis E. coli (NMEC) and avian pathogenic E. coli (APEC) [1,2].
The E. coli genome exhibits extensive plasticity, with ~55% gene conservation observed across strains from different pathotypes [3,4]. Moreover, pathogenic E. coli strains have a more diverse gene content than their commensal counterparts [3,5]. Even in a single multidrug resistant ExPEC lineage such as the globally disseminated ST131 clone, the core genome accounts for only ~75% of the average genome size [6,7]. Studies on the evolution of ST131 showed that the lineage is comprised of three distinct clades, namely clade A, clade B and clade C. Strains from Clade C form a globally dominant fluoroquinolone resistant group, and can be further divided into two subclades (C1 and C2), of which strains from C2 are associated with the presence of the extended spectrum beta-lactamase CTX-M-15 gene [6–12].
The accessory genome, defined as genomic elements present in only a subset of E. coli strains, includes prophages, plasmids, integrative and conjugative elements (ICEs), and genomic islands (GIs); these elements play an important role in E. coli evolution, adaptation, resistance and virulence [13–21]. The most striking difference between genomes of non-pathogenic and pathogenic E. coli lies in GIs, which are often >100 kb in size and generally contain multiple virulence and/or fitness genes acquired through lateral gene transfer [3,5,14,22,23]. GIs possess various combinations of the following features: (i) a G+C content different from the average G+C content of the core genome, (ii) a modular arrangement of smaller mobile elements, (iii) differing levels of stability, (iv) a strong association with pathogenic strains, (v) integration sites near tRNA genes, and (vi) multiple virulence genes (also referred to as pathogenicity islands—PAIs) or resistance genes (associated with resistance islands) [16,24–28]. In E. coli, hotspots for the integration of GIs are usually located at the selC-tRNA, leuX-tRNA, thrW-tRNA, serT-tRNA, serW-tRNA, asnT-tRNA, pheV-tRNA and pheU-tRNA [23]. Among these sites, pheV-tRNA represents the major integration hotspot for GIs containing virulence factors associated with pathogenic E. coli such as UPEC, EHEC, EPEC and ETEC [23,29–31]. For example, the GI-pheV found in CFT073 (denoted CFT073_GI-pheV) is found more frequently in strains associated with acute pyelonephritis and cystitis than in fecal strains [29]. In addition, GIs containing the locus of enterocyte effacement (LEE), which encodes a type III secretion system, are inserted at pheV-tRNA in many EHEC and EPEC strains [18,32]. Finally, pheV-tRNA (and pheU-tRNA) are also insertion hotspots for ICEs such as ICEEc2, which can comprise both virulence and antibiotic resistance genes [33].
Despite the wealth of knowledge on E. coli genomics, the study of GIs and their evolution has been challenging due to their large size, diverse gene content, highly mosaic structure and fragmented assembly in draft genomes [34]. Here, we examined the GI integrated at pheV-tRNA (GI-pheV) to understand its contribution to E. coli evolution and virulence, with a focus on its evolution in the ST131 lineage.
Results
The presence of GI-pheV is associated with pathogenic E. coli
We began by determining the prevalence of GI-pheV in 2,328 E. coli complete genomes from NCBI Reference Sequence Database (RefSeq, retrieved on 23/12/2022) based on the presence of pheV-tRNA, the integrase encoding gene intP4 and a coding sequence located upstream of the pheV-tRNA. Based on these criteria, GI-pheV was identified in 52% (1,242/2,383) E. coli complete genomes (S1 Table). However, the prevalence of GI-pheV was different among E. coli phylogroups and sequence types (STs; Figs 1 and S1). GI-pheV was most prevalent in phylogroups associated with pathogenesis including ExPEC-associated phylogroups B2 (82%), D (71%) and F (63%), and EHEC-associated phylogroup E (92%). Furthermore, GI-pheV was present in >90% of strains from major ExPEC STs such as ST131, ST73 and ST127, and Shiga toxin-producing O157:H7 E. coli ST11 (S1 Fig). In contrast, GI-pheV was present in <40% of non-pathogenic phylogroups A (34%) and B1 (39%).
The total number of E. coli genomes in each phylogroup is shown in brackets.
Next, we established a reference set of 66 completely sequenced E. coli genomes representing a range of phylogroups and STs available on the NCBI Genbank database, and comprising non-pathogenic (n = 23) and pathogenic (n = 35) strains (S2 Table). In total, GI-pheV was detected in 41/66 (62%) genomes from this dataset. The location, size and direct repeat sequences for the 41 identified GI-pheV are listed in Table 1. GI-pheV was found in all pathogenic strains, but only in 26% (6/23) of non-pathogenic strains, indicating a strong association of GI-pheV and pathogenic E. coli (P < 0.0001, Chi-square test; S2 Table). GI-pheV was present in all phylogroup E (n = 12), phylogroup D (n = 2) and phylogroup F (n = 3) strains, as well as 8/15 (53%) phylogroup B1 strains and 12/19 (63%) B2 strains. In contrast, GI-pheV was only identified in 4/15 (27%) phylogroup A strains, which comprise non-pathogenic strains. The size of the 41 GI-pheV was extremely variable, ranging from 9.7 kb (11368_GI-pheV) to 110 kb (55989_GI-pheV) (Table 1). Among the 41 identified GI-pheV, only 27 (66%) contained direct repeats flanking the GI (Table 1). These sequences upstream (attL) and downstream (attR) of the GI-pheV sequences were mostly either a 23-bp direct repeat (16 strains) or a 19-bp direct repeat (10 strains), and include part of the pheV-tRNA gene sequence (Table 1). GI-pheV that lacked a recognisable attR (mean 23.5 kb, Q1 –Q3 18 kb– 32.1 kb) were significantly smaller than those where an attR (mean 75 kb, Q1 –Q3 52.1 kb– 87.4 kb) was identified (P < 0.0001, unpaired t-test).
GI-pheV can be classified by the type of intP4 integrase gene
Core-genome analysis of the 41 GI-pheV sequences from our dataset revealed that the only conserved coding sequence in all completely sequenced GI-pheV was the intP4 gene (S2 Fig). The intP4 gene encodes a P4 integrase type tyrosine site-specific recombinase and is located immediately adjacent to the pheV-tRNA. Therefore, to examine the evolution of GI-pheV, we analyzed the phylogenetic relationship of the intP4 gene from the 41 GI-pheV described above. The most divergent intP4 genes were identified in GI-pheV from the environmental E. coli strain SMS-3-5 and the commensal E. coli strain REL606, sharing approximately 46.3% and 85% nucleotide conservation with intP4 in other GI-pheV, respectively. In contrast, the 39 intP4 genes from the remaining GI-pheV were very closely related (95% of intP4 with ≥ 94% nucleotide sequence conservation). Of note, we observed a similarly high sequence conservation of intP4 among the 1,242 GI-pheV-positive E. coli complete genomes in the larger RefSeq database; 98% of intP4 genes shared > 94% nucleotide sequence conservation (S3 Fig), indicating that the analysis of 39 intP4 genes was sufficient to capture major hallmarks of GI-pheV variation in E. coli.
Phylogenetic analysis revealed that most of 39 intP4 genes of GI-pheV could be clustered into 5 main alleles, that we have named intP4.1 to intP4.5 (Fig 2). The nucleotide sequence of each intP4 allele was highly similar (< 1% divergence), but divergence between intP4 alleles ranged from 4–6%. The intP4 alleles displayed a phylogroup association; most alleles in phylogroup B1 strains belonged to intP4.5, most alleles in phylogroup B2 strains belonged to intP4.1 or intP4.3, and most alleles in phylogroup E strains belonged to intP4.2 or intP4.4. The intP4 alleles were also largely congruent with strain pathotypes; e.g. intP4 from InPEC strains comprised intP4.2, intP4.4 and intP4.5 alleles, while intP4 from ExPEC strains comprised intP4.1 and intP4.3 alleles (Fig 2).
A phylogenetic tree of 40 intP4 nucleotide sequences (except SMS-3-5_intP4) was constructed using RaxML, with 1,000 bootstraps, and the scale as the number of SNPs. The intP4 gene clustered into 5 main alleles, namely intP4.1 –intP4.5 (coloured across each group for clarity). Highly supported branches are marked with one (*) or two (**) asterisks based on its bootstrapped value (>50% and >90%, respectively). E. coli phylogroups, pathotypes and the 3’ end direct repeat (attR) of GI-pheV are shown on the right of the intP4 phylogeny (indicated by black boxes). Strain types: non-path, non-pathogenic; EPEC, Enteropathogenic E. coli; EHEC, Enterohemorrhagic E. coli; ETEC, Enterotoxigenic E. coli; EAEC, Enteroaggregative E. coli; UPEC, Uropathogenic E. coli; NMEC, neonatal meningitis associated E. coli; APEC, Avian pathogenic E. coli.
GI-pheV gene content is associated with intP4 integrase type
Given the role of the P4 integrase in GI-pheV mobilization, we hypothesized that GI-pheV containing the same intP4 allele would have the same or similar gene content. In fact, GI-pheV harbouring different intP4 groups have different gene content (S4 Fig). Comparative analyses of intP4.1-harbouring GI-pheV (hereafter referred as intP4.1_GI-pheV) revealed that they possessed very similar gene content, irrespective of the orientation and location of the genes in the GI (Fig 2). Similar to other GIs, intP4.1_GI-pheV has a mosaic structure comprising several gene modules, which we define as clustered coding sequences within a GI separated by insertion sequences (Fig 3). The intP4.1_GI-pheV contained genes encoding various fitness factors and/or virulence factors associated with uropathogenesis, including a complete or partial pyelonephritis-associated P fimbriae gene cluster (pap), the secreted autotransporter serine protease gene (sat), an adhesin gene (iha), the aggregation- and biofilm-associated antigen 43 gene (flu), the aerobactin siderophore biosynthesis gene cluster (iucABCD) and the aerobactin receptor (iutA) (Fig 3). GI-pheV from six out of the eight UPEC genomes examined that contained intP4 belonged to intP4.1. While a full pap operon is present in CFT073_GI-pheV and UMN026_GI-pheV, other GI-pheV in the UPEC strains examined, including EC958_GI-pheV and JJ1886_GI-pheV, only harboured remnants of the pap locus. All intP4.1_GI-pheV except CE10_GI-pheV contained a 19-bp imperfect direct repeat (with a nucleotide substitution) at the 3’ of the island (Table 1). CE10_GI-pheV was the smallest intP4.1_GI-pheV examined (46 kb) and contained the fewest virulence-associated genes (Fig 3).
Gene content of intP4.1_GI-pheV visualized based on an alignment of the entire GI using Easyfig (drawn to scale). The black to grey gradient between the GI indicates their pairwise nucleotide conservation. Conserved genes and gene modules are colour coded as indicated in the text. Insertion sequences (transposons, transposases and other mobile elements) are drawn as black arrows.
The intP4.2_GI-pheV was identified in strains belonging to three different phylogenetic groups: A, B2 and E (Fig 1). The core component of intP4.2_GI-pheV was restricted to the fec cluster (which encodes a ferric citrate transport system). The remaining gene content of intP4.2_GI-pheV was variable, with a remnant of the yee module (which contains coding sequences related to the CP4-44 prophage) present in H10407_GI-pheV and ED1a_GI-pheV, and a remnant of the flu module present in H10407_GI-pheV (S4 and S5 Figs).
The intP4.3_GI-pheV contains a number of genes including tia (encoding an invasion determinant of ETEC), a full pap operon with the same gene composition as the one in CFT073_GI-pheV but with extensive sequence variation in the papA gene encoding the major pilin protein PapA (67% amino acid sequence identity) and the papX gene encoding the PapX regulator (22% amino acid sequence identity), pgt (encoding a phosphoglycerate transport protein), ireA (encoding an iron-responsive element), ibr (encoding an immunoglobulin-binding protein) and a remnant of the yee module (S4 and S5 Figs). The ibr and yee modules were also identified in intP4.5_GI-pheV (S4 Fig). A typical intP4.5_GI-pheV, for example 55989_GI-pheV, contained a gene encoding a putative DNA helicase, a gene encoding a serine protease (pic), and a gene cluster encoding a putative type VI secretion system (T6SS) (S4 and S5 Figs).
The gene content of intP4.4_GI-pheV was highly conserved in the strains examined. Strains that possessed intP4.4_GI-pheV belonged to phylogroup E and were either EPEC or EHEC. The intP4.4_GI-pheV shared the same gene content as the previously described Sakai_GI-pheV (SpLE3) (also known as OI-122 in EDL933) [35–37], which contains EHEC-associated virulence factors including an adherence/lymphocyte inhibition factor (efa1/lifA), an enterotoxin homologous to the Shigella flexneri SenA (ent) and a virulence factor associated with Salmonella typhimurium survival in macrophages (pagC) (S4 and S5 Figs). Most intP4.4_GI-pheV did not contain direct repeat (attR) sequences, the exceptions being 12009_GI-pheV, CB9615_GI-pheV and RM12579_GI-pheV, which possessed a 23-bp imperfect attR (Fig 1). These GI-pheV were larger than the other intP4.4_GI-pheV, and contained remnants of the flu and yee modules in addition to components of Sakai_GI-pheV. In strain 12009, the core LEE (encoding T3SS effectors and structure proteins—esp, sep, ces, and the translocated intimin receptor—tir) is located downstream of the 12009_GI-pheV attR direct repeat, suggesting its independent acquisition in this strain.
Gene modules and individual genes of intP4.1_GI-pheV have evolved differently
The analyses described above demonstrate that intP4.1_GI-pheV is largely restricted to UPEC. Indeed, analysis of the 1,242 GI-pheV-positive E. coli complete genomes in the RefSeq database revealed the intP4.1 allele was predominantly found in UPEC/ExPEC associated phylogroups B2 (57%), D (49%) and F (42%) compared to phylogroup A (6%) (S1 Table). Thus, the remainder of this study focused on this GI to understand its evolution and contribution to UPEC virulence. Phylogenetic analysis of nine intP4.1_GI-pheV nucleotide sequences showed that intP4.1_GI-pheV could be divided into two well-supported clusters, of which cluster 1 comprised CFT073_GI-pheV and UMN026_GI-pheV and cluster 2 comprised the remaining seven GI-pheV (Fig 4A). The integrase gene of CFT073_GI-pheV and UMN026_GI-pheV were the most divergent within the intP4.1 group, differing from other strains by 13 SNPs and 6 SNPs, respectively. Given the comparative small size of CE10_GI-pheV, we removed it from our dataset and repeated the analysis. The resulting investigation based on a 64,922-bp core GI-pheV produced a similar overall phylogeny, with the exception that EC958_GI-pheV segregated separately from the other GIs (Fig 4B).
Whole GI-pheV or gene modules were aligned using Mauve or ClustalO, respectively. Phylogenetic trees were constructed using RAxML, with 1,000 bootstraps. Highly supported branches are shown with bootstrapped values and scales indicating the number of SNPs. Phylogenetic trees depicted as follows: (A) nine intP4.1_GI-pheV (phylogenetic tree constructed based on a 30,859 bp—core GI-pheV). (B) eight intP4.1_GI-pheV (phylogenetic tree excluding CE10_GI-pheV, constructed based on a 64,922-bp—core GI-pheV). (C) shiF_iucABCD_iutA_sat module. (D) nan module. (E) flu module. (F) yee module.
The intP4.1_GI-pheV possesses a mosaic structure with several modules separated by insertion sequences (Fig 3). To examine the coevolution of these elements, we analyzed the phylogenetic relationship between each module, designated as: (i) the shiF_iucABCD_iutA_sat module (14,133 bp), (ii) the nan module (8,905 bp), (iii) the flu module (11,415 bp), and (iv) the yee module (8,951 bp) (Fig 3). Overall, the modules from CFT073_GI-pheV and UMN026_GI-pheV clustered together with the exception of the flu module, where UMN026_flu possessed a closer relationship to EC958_flu (Fig 4). The nan module showed the highest conservation (99.7% nucleotide sequence conservation), while other modules are more diverse, with nucleotide sequence identity ranging from 95%–97% (Fig 4).
Next, we investigated the genetic variation between individual genes within intP4.1_GI-pheV. Nucleotide sequence conservation varied among the individual genes examined, despite their preserved modular location (Fig 5). The major contributor to nucleotide diversity was the iron-acquisition genes from CFT073 and UMN026, suggesting they have evolved under strain-specific selective pressures (Fig 5). In the shiF-iucABCD-iutA-sat module, although shiF, iucA, iucB and sat share more than 99.5% nucleotide sequence identity, iucC, iucD and iutA from CFT073_GI-pheV and UMN026_GI-pheV are comparatively diverse (98%, 96% and 89% nucleotide sequence identity compared to their corresponding genes in other intP4.1_GI-pheV, respectively) (Fig 5). Similarly, the SNP frequencies of individual genes (determined by the number of SNPs over the entire gene length) varied considerably across GI-pheV (S3 Table). While shiF, iucA, iucB and sat had 5.8–7.4 SNPs/Kb, the SNP frequencies were approximately 5 and 17 times higher in iucD and iutA, respectively. Although previous reports have found diversifying (positive) selection in ferrisiderophore receptors from other organisms such as Pseudomonas aeruginosa [38], pairwise dN/dS analysis between CFT073 and EC958 suggests that most genes in intP4.1_GI-pheV are under negative (purifying) selection (dN/dS < 1) (Fig 5). The most likely explanation for higher SNP frequencies in iucD and iutA is by recombination of a region overlapping both iucD and iutA prior to acquisition of the shiF-iucABCD-iutA-sat module by the last common ancestor of CFT073_GI-pheV and UMN026_GI-pheV. Overall, this analysis supports the hypothesis that individual modules of intP4.1_GI-pheV have evolved independently via mutation and/or recombination.
The average dN/dS ratio is indicated with a triangle (value range on the left y-axis). Nucleotide similarity, compared to the consensus sequences, is shown with box plots (value range on the right y-axis), with the middle line showing the group median and the coloured dots as the group outliers (defined as any value outside 1.5 times the interquartile range below the first quartile).
A specific GI-pheV is common to ST131
Comparative genomic analysis revealed that GI-pheV in two completely sequenced E. coli ST131 clade C2 strains, EC958 (isolated in the UK [39]) and JJ1886 (isolated in the USA [40]), have the same gene content (Fig 3). To examine the gene content of GI-pheV more broadly in ST131, we used a previously published dataset comprising 102 genomes from six different geographic regions that accurately reflects the diversity and clade structure of this lineage [6,7]. A blastn search of these genomes revealed EC958_GI-pheV is present in 91 strains (89%), with 82 strains possessing a GI-pheV almost identical to EC958_GI-pheV (hereafter referred to as ST131_GI-pheV) (Fig 6). This high level of ST131_GI-pheV conservation occurred independent of clade phylogeny. In strains that did possess differences in ST131_GI-pheV, these were associated with a lack of one or more gene modules. Differences were observed in some strains from each clade. For example, in clade A strains, the shiF-iucABCD-iutA-sat module was absent from S31EC and S94EC, consistent with a single deletion event within their respective GI-pheV islands. Similarly, the nan, flu, and yee modules were absent from the clade C strain P50EC (Fig 6). Strain S39EC contains a remnant of ST131_GI-pheV, with only intP4 and shiA remaining. The deletion of whole modules, but not individual genes, would be consistent with a single deletion event mediated by flanking IS sequences.
The gene content of EC958_GI-pheV is shown along the x-axis, with the annotation on top (drawn to scale). Strain identifiers are on the y-axis together with the ST131 phylogenetic tree (SNP based), coloured according to clade designations: red, clade A; orange, clade B; pink, B-C intermediate clade; green, clade C. The presence of ST131_GI-pheV in ST131 was identified by BLASTn analysis of ST131 genome assemblies using EC958_GI-pheV as a reference and a cut-off at 95% nucleotide conservation. The BLASTn data was visualized with SeqFindr. The presence of GI-pheV and the type of intP4 and attR are indicated on the right.
We next investigated the intP4 gene sequences in the ST131_GI-pheV islands. Of 91 ST131_GI-pheV islands, we excluded 12 that lacked a full length intP4 gene, presumably due to IS. Most of the remaining 79 ST131_GI-pheV islands harboured an identical intP4.1 allele and the same 19-bp imperfect right direct repeat (attR) sequence as in either JJ1886_attR or EC958_attR (Fig 6). The fact that most GI-pheV in ST131 harbour very similar gene content, in addition to the conserved nature of their cognate intP4 and attR sequences, suggests that the acquisition of ST131_GI-pheV into ST131 may have occurred in the common ancestor of ST131. Subsequent to this event, the deletion of individual modules or even the entire GI appears to have occurred sporadically in a strain-specific manner.
ST131_GI-pheV has co-evolved with the ST131 core genome
ST131_GI-pheV is the predominant GI-pheV type in ST131, and is found across all three ST131 clades. To better understand the evolution of ST131_GI-pheV, we analyzed their phylogenetic relationship. Excluding the 12 ST131_GI-pheV that lacked a full-length intP4 gene and 5 that had gene module deletions (S131EC, S94EC, G216, P50EC and S39EC) (Fig 6), we generated a 47,550-bp core ST131_GI-pheV alignment. Phylogenetic analysis of this core alignment showed that ST131_GI-pheV formed three well-supported groups, ST131_GI-pheV.1, ST131_GI-pheV.2 and ST131_GI-pheV.3 (Fig 7A). ST131_GI-pheV.1 are mostly found in clade A strains whereas ST131_GI-pheV.3 is found across clades B and C. ST131_GI-pheV.2 includes EC958 and 11 closely related clade C2 strains as well as a well-supported sister group of clade B strains. The fact that the ST131_GI-pheV phylogeny is mostly congruent with the core genome phylogeny strongly suggests this element has co-evolved with its corresponding core genome.
Whole GI-pheV or gene modules were aligned using Mauve or ClustalO, respectively. Phylogenetic trees were constructed using RAxML, with 1,000 bootstraps. Highly supported branches are shown as asterisks for a bootstrapped value > 75%, with scales indicating the number of substitution SNPs. Branches are coloured according to the ST131 clades to which the majority of strains belong: red, clade A; orange, clade B; pink, B-C intermediate clade; green, clade C. Individual strains that cluster with strains from different clades are indicated by filled circles or strain names coloured with their corresponding clade. (A) EC958_GI-pheV. (B) shiF_iucABCD_iutA_sat module. (C) nan module. (D) flu module. (E) yee module.
The phylogenetic discrepancies apparent in ST131_GI-pheV.2 compared to ST131 phylogeny prompted us to investigate whether there were any different evolutionary trajectories for components of GI-pheV across ST131. Similar to what we observed in intP4.1_GI-pheV from the E. coli complete genomes, there were also differences in the SNP distribution of gene modules in ST131_GI-pheV (Fig 7). The shiF-iucABCD-iutA-sat module is extremely well conserved in ST131_GI-pheV, with fewer than 10 SNPs over a 13,889 bp-alignment, mostly in strain CD306 (Fig 7B). Similarly, the nan module was conserved in ST131_GI-pheV except in strains carrying ST131_GI-pheV.2, suggesting that recombination within a hypothetical gene in the nan module (locus tag EC958_3299) is partly responsible for the phylogenetic placement of ST131_GI-pheV.2 (Fig 7C). In contrast, we observed a high SNP density in both the flu and yee modules, with 494 SNPs over 5,707-bp alignment and 142 SNPs over 2,510-bp alignment, respectively (Fig 7D and 7E). In addition, the flu module from a subset of 12 clade C strains containing ST131_GI-pheV.2 formed a distinct group, suggesting recombination has shaped its composition in these strains. Indeed, the SNP frequencies in the flu and yee modules are 8.66 x 10−2 and 5.66 x 10−2 SNPs per base pair, which are higher than the average SNP density in ST131 recombination regions (1.19 x 10−2) and significantly higher than that in the non-recombination regions reported previously for ST131 (1.39 x 10−3) [6]. Together with the nan module, these differences distinguish ST131_GI-pheV.2 from other ST131_GI-pheV. In all modules, increased SNP frequency was observed in several strains (e.g. CD306, S65EC, S22EC and JJ2244), providing evidence for strain-specific evolution of GI-pheV.
Other GI-pheV types in E. coli ST131
We next investigated the presence of GI-pheV in 11 strains that did not contain ST131_GI-pheV. The presence of GI-pheV was manually determined based on the presence of the intP4 gene immediately after pheV-tRNA. Among the 11 strains, four strains (HVM277, HVM52—clade B, H17—intermediate B-C clade and S124EC—clade C) did not harbour intP4, suggesting the deletion of the entire island. The intP4 gene was found in the remaining seven strains (S2EC, S26EC and S37EC—clade A; HVM2044, HVM2289, S79EC and S128EC—clade B), indicating the presence of a GI-pheV with different gene content compared to ST131_GI-pheV. The phylogenetic relationship of intP4 from these seven GI-pheV revealed they differ from intP4 in EC958_GI-pheV, with the nucleotide divergence ranging from 2.6%–10.3%. Compared to intP4 from the 66 complete E. coli genome dataset, intP4 in S2EC, S26EC and S37EC (clade A) clustered in the same group as of REL606_intP4, while intP4 in S79EC and S128EC (clade B) formed a subgroup with 536_intP4 (Fig 8). Indeed, these two intP4 genes differ from the intP4 gene of the pyelonephritis strain 536 by only two SNPs. The intP4 gene in HVM2044 and HVM2289 was closest to the intP4.1 allele group, but formed a separate well-supported clade (Fig 8).
A phylogenetic tree of ST131_intP4 together with 40 intP4 from 66 complete E. coli genomes examined in this study was constructed using RaxML, with 1,000 bootstraps. The majority of the ST131_intP4 clustered in the intP4.1 allele, together with other UPEC strains. ST131 strains are indicated by coloured dots in accordance with their clade designation: red, clade A; orange, clade B; pink, B-C intermediate clade; green, clade C. Highly supported branches are marked by an asterisk (bootstrapped value >75%).
Due to the presence of insertion sequences and repetitive regions within GI-pheV, we could not determine the complete gene content of GI-pheV in the above seven strains using de novo assemblies from Illumina sequencing alone. To overcome this problem, the complete genome of one representative strain from each different intP4 (S37EC, S79EC and HVM2044) was determined by PacBio sequencing. Analysis of these genomes revealed the complete GI-pheV sequences shared little similarity with ST131_GI-pheV; however, their gene content was similar to other GI-pheV of the same intP4 type (Fig 9). For example, S37EC_GI-pheV and REL606_GI-pheV shared 97% nucleotide sequence conservation in their respective flu/yee modules (Fig 9). Similarly, S79EC_GI-pheV and 536_GI-pheV both contain near identical pix gene clusters (encoding Pix fimbriae) and pgtABCP loci (encoding a phosphoglycerate transport system) downstream of their respective intP4 allele. In addition, S79EC_GI-pheV had the same 23-bp imperfect right direct repeat as in 536_GI-pheV, which is different from that found in ST131_GI-pheV. Although HVM2044_intP4 was closest to the intP4.1 allele found in ST131_GI-pheV (Fig 8), the gene content of HVM2044_GI-pheV was substantially different (Fig 9). HVM2044_GI-pheV contained a complete pap gene locus (P fimbriae), a gene encoding a DNA methylase, and several regions not found in ST131_GI-pheV, as well as a 23-bp imperfect attR. Together, these data suggest that even in a single ST131 lineage, several different types of GI-pheV can be found. However, ST131_GI-pheV remains the predominant type of GI-pheV in ST131.
Various gene content of GI-pheV visualized based on the alignment of the entire GI using Easyfig. The black to grey gradient between the GI indicates their pairwise nucleotide conservation. Conserved genes and gene modules are coloured as indicated in the text. Strain names are coloured in accordance with their ST131 clade designation: red, clade A; orange, clade B; green, clade C.
Discussion
It is well established that GIs contribute to the evolution and pathogenesis of E. coli, but their large size and genetic diversity makes the study of their evolutionary trajectory challenging. Here we examined the evolution of GI-pheV, a GI integrated at pheV-tRNA on the E. coli chromosome.
We found that the presence of GI-pheV is significantly associated with pathogenic E. coli compared to non-pathogenic strains. In addition, GI-pheV prevalence is higher in phylogroups associated with pathogenesis compared to commensalism. In agreement with previous studies of E. coli GIs [23,41], GI-pheV was shown to have a mosaic structure containing gene modules flanked by insertion sequences and/or mobile elements. The gene content of GI-pheV is extremely diverse, with the P4-like integrase encoding gene intP4 the only conserved gene. It has previously been shown that the GI-encoded integrases of E. coli are related to the integrase of phage P4 [34]. Our analyses identified five intP4 alleles, the majority of which grouped in concordance with the whole genome phylogeny, suggesting co-evolution of each GI-pheV type with the E. coli chromosome. In addition to intP4, the flu module and partial yee module were found on different GI-pheV. These modules are also present on GIs from other species, including the pathogenicity island in Shigella flexneri and the cryptic prophage CP4-44 in E. coli K-12 [42,43]. Together, this suggests that the original acquisition of GI-pheV by E. coli could have been via bacteriophage integration at the pheV-tRNA locus, with subsequent events that have shaped this element including the loss of bacteriophage genes, recombination, and acquisition and deletion of gene modules.
The length and the composition of direct repeats that flank GIs are important for their mobility, and thus the integrity of these elements impacts the stability of GIs on the chromosome [44,45]. For example, a single nucleotide mismatch in the direct repeats can reduce the frequency of GI deletion by several orders of magnitude [46]. The effect on stabilization of GIs into the host chromosome is exemplified by the high-pathogenicity island (HPI) of Yersinia enterocolitica, which lacks intact flanking repeats and is stably maintained on the chromosome in the species [47]. We showed that GI-pheV that lack direct repeats are shorter than those with direct repeats likely due to ancestral deletions that removed the attR sites and stabilization into the host chromosome. In intP4.1_GI-pheV, most GIs contain a 19-bp imperfect repeat, of which attR has a T substitution compared to the attL repeat. The attR from EC958_GI-pheV also contains a second T substitution that is also found in ST131 strains that carried ST131_GI-pheV.2 and may be associated with increased stability compared to other intP4.1_GI-pheV.
Our analysis of GI-pheV in UPEC revealed that while its overall gene content is conserved, modules and individual genes exhibit variation suggestive of strain specific evolution. A similar phenomenon has been reported for the LEE of atypical EPEC [18]. The flu and yee genes are diverse, likely due to recombination, with only one gene showing evidence for positive selection in UMN026. In contrast, genes involved in sialic acid catabolism (nan module) are highly conserved (99.6% nucleotide conservation). Sialic acid is abundant in human mucus and can be utilized by E. coli as a sole source of carbon and energy. Sialic acids also contribute to the regulation of type 1 fimbriae, providing a mechanism to balance pathogen-host interactions [48]. Together, the conserved nan modules among GI-pheV in UPEC may enhance its capacity to colonise host epithelial surfaces, such as in the intestinal and urinary tract. Similarly, the iucABCD genes encoding the aerobactin system involved in iron acquisition are largely conserved and may contribute to UPEC virulence. However, a divergent region of sequence overlaps both the ferric aerobactin receptor gene iutA and the upstream iucD gene in CFT073 and UMN026 compared to other strains that contain the intP4.1_GI-pheV. The phylogenetic distance between CFT073 and UMN026 and the high nucleotide identity shared by their respective iucD/iutA regions (and, indeed, the rest of their respective GI-pheV sequences) are consistent with recombination and subsequent lateral gene transfer of GI-pheV. Although we did not find any evidence for positive selection in iutA, our data suggest that it may have evolved independently within this module in some UPEC strains. Polymorphisms in iutA have also been observed in other E. coli pathotypes, Shigella spp., Klebsiella pneumoniae and Vibrio hollisae [49–51].
ST131 was initially identified as a clone associated with the production of CTX-M-15 [52–54]. Since then, ST131 has become recognised globally as one of the most successful ExPEC lineages [6,7,10,55,56]. Here we investigated GI-pheV in a previously validated ST131 genome dataset, and revealed dynamic changes in this element over a short evolutionary timespan. GI-pheV was present in 96% of the ST131 isolates investigated, most of which contained a GI-pheV similar to that found in the reference ST131 strain EC958. These ST131_GI-pheV islands possess a near-identical intP4.1 allele (nucleotide divergence of 1–2 nonsynonymous SNPs) and highly conserved 19-bp direct repeats. Furthermore, the ST131_GI-pheV islands were found across all ST131 clades, including a clade B-C intermediate strain JJ1897.
Different levels of diversity were observed among the GI-pheV gene modules in the ST131 strains examined. The iucABCD-iutA and nan module displayed very high conservation, while the flu and yee modules were the most variable. Although the ST131_GI-pheV was the most common, an additional three different types of GI-pheV were also identified in ST131. These GI-pheV possess distinct types of intP4 and direct repeats, consistent with their independent acquisition. The strains that harbour these alternate GI-pheV generally clustered together with other strains that lack a GI-pheV altogether. For example, HVM2044 and HVM2289, which have a GI-pheV similar to 536_GI-pheV, clustered with HVM277 and HVM52, which do not have GI-pheV. Based on these data, we propose the following evolutionary trajectory of GI-pheV in ST131: (i) the acquisition of ST131_GI-pheV in the common ancestor of ST131; (ii) co-evolution of ST131_GI-pheV in the three major ST131 clades, including recombination within gene modules; (iii) sporadic loss of gene modules or the entire GI; and (iv) the acquisition of other GI-pheV types by some strains that had previously lost the ST131_GI-pheV.
In summary, this work presents a comprehensive evolutionary study of the GI-pheV element. While the GI-pheV gene content exhibits extensive diversity, we observed that this follows a pathotype-specific association. In UPEC, GI-pheV contains a number of virulence and fitness modules linked to colonization of the urinary tract. The evolution of GI-pheV in ST131 is more complex, with our analysis revealing it was acquired early in the evolution of ST131, with loss or gain of genes and modules via recombination shaping its composition.
Materials and methods
E. coli strains and genomic data
The 2,382 E. coli complete genomes were retrieved from NCBI Reference Sequence Database on 23/12/2022 (S1 Table). The reference set of 66 completely sequenced E. coli genomes comprising 23 non-pathogenic and 43 pathogenic strains (28 intestinal and 15 extra-intestinal) was downloaded from Genbank (S2 Table). In silico multi-locus sequence typing (MLST) was performed according to the seven-gene scheme [57]. Read data and de novo assemblies of 95 E. coli ST131 were obtained from previous studies [6,7]. The genome sequences of an additional seven strains [58] that represented an intermediate link between ST131 strains from clades B and C [7] were also analysed. De novo assemblies were ordered against the genome of the reference ST131 strain EC958 [59] using Mauve [60].
Three ST131 strains (HVM2044, S79EC and S37) that harboured GI-pheV and exhibited extensive diversity compared to the EC958 ST131_GI-pheV were sequenced on a PacBio RS II sequencing instrument using 3 SMRT cells, using a 10-kb insert library and the P6-C4 sequencing chemistry. De novo genome assembly was performed using PacBio’s SMRT Portal (v2) and the hierarchical genome assembly process (HGAP v2.0) with default settings and a seed read cut-off length of 5 kb. The complete genome sequences were annotated using Prokka [61] and insertion sequence (IS) annotation was done with ISFinder (https://www-is.biotoul.fr/). Annotation of CDS and IS were then curated manually, with a focus on GI-pheV. PacBio reads were deposited in the SRA with the following accession numbers: SRR10761087, SRR10761088 and SRR10761089.
GI-pheV identification and comparative genomic analysis
The prevalence of GI-pheV in the 2,382 E. coli complete genomes was evaluated based on the presence of the gene located upstream of pheV-tRNA, pheV-tRNA and the integrase intP4 gene. The 41 GI-pheV from the reference set of 66 E. coli complete genomes were manually identified based on the presence of an integrase encoding gene at the 3’ end of the pheV-tRNA. The 3’ end of GI-pheV was determined based on the presence of a direct repeat sequence (attR) that represents the remnant of pheV-tRNA. Where a direct repeat was not identified, an approximation of the 3’ end of GI-pheV was determined based on comparison to the genome of the E. coli K-12 strain MG1655, and was determined as the last nucleotide before the conserved region found in the same location in MG1655 or at the breakpoint before the kps locus. The gene content of each GI-pheV was compared using Mauve, ACT and EasyFig [60,62,63].
The prevalence of ST131_GI-pheV in the sequenced ST131 strain set was examined using BLASTn [64] and visualized with SeqFindr (https://github.com/mscook/seqfindr). The presence/absence of GI-pheV was presented to scale by performing analyses over a 200 bp sliding window, applying a nucleotide identity cut-off at 0.95. The locations of GI-pheV in the E. coli ST131 draft genomes were identified manually. Briefly, the contigs from de novo assemblies of 102 E. coli ST131 strains were ordered against the complete EC958 genome. GI-pheV was then annotated based on the presence of the intP4 gene located downstream of pheV-tRNA. The end of the GI-pheV was defined based on pairwise comparison to EC958_GI-pheV. The direct repeat attR was identified by manually searching for attR sequences listed in Table 1.
Sequence alignment and phylogenetic analysis
GI-pheV nucleotide sequences from complete genomes were aligned using Mauve [60]. Regions larger than 100 bp were extracted from the alignment and concatenated using the stripSubsetLCBs script supplied by Mauve to generate the core GI-pheV alignment. Individual genes and regions of interest were aligned using ClustalO and visualized using SeaView [65,66]. The ratio of non-synonymous to synonymous substitutions (dN/dS ratio) was calculated with SNAP [67,68] using codon-aligned nucleotide sequences of individual genes generated by SeaView program [66]. The dN/dS ratio is used to indicate if a gene is conserved or may have undergone adaptive evolution, with dN/dS > 1 indicating positive (adaptive or diversifying) selection, dN/dS = 1 indicating neutral evolution, and dN/dS < 1 indicating negative (purifying) selection. Phylogenetic trees of the core GI-pheV alignments and other alignments were constructed with RaxML v.7.2.8 using substitution-only single nucleotide polymorphisms (SNPs) with the general time reversible (GTR) GAMMA model of among site rate variation (ASRV) [69]. The robustness of the trees was tested using 1,000 bootstraps. Trees were visualized and edited using FigTree v1.3.1 (http://tree.bio.ed.ac.uk/software/figtree/) as previously described [70–72].
Supporting information
S1 Fig. Prevalence of GI-pheV in STs represented in the 2,382 E. coli complete genomes from RefSeq Database.
Bars are coloured according to the corresponding phylogroup. The total number of E. coli genomes in each ST is shown in brackets.
https://doi.org/10.1371/journal.pgen.1011459.s001
(TIF)
S2 Fig. Pangenome of GI-pheV.
Gene presence and absence of GI-pheV from 40 E. coli complete genomes, together with their GI-pheV-intP4 phylogeny.
https://doi.org/10.1371/journal.pgen.1011459.s002
(TIF)
S3 Fig. Sequence conservation of intP4 in GI-pheV from the 1,242 GI-pheV-positive E. coli complete genomes.
Shown is the percentage nucleotide identity of intP4 in GI-pheV from different phylogroups (A) and STs (B).
https://doi.org/10.1371/journal.pgen.1011459.s003
(TIF)
S4 Fig. Comparison of gene content in representative GI-pheV from different intP4 alleles.
Nucleotide comparison of a representative GI-pheV from each intP4 allele revealed extensive diverse, with intP4 as the only conserved core gene. The outer ring was coloured according to intP4 alleles as in Fig 1 (blue, intP4.1; green, intP4.2; violet, intP4.3, light orange, intP4.4 and yellow, intP4.5). Inner rings are regions on each GI-pheV coloured according to the following scheme: blue, virulence/fitness factors; orange, insertion sequences and mobile elements; light grey, hypothetical proteins. Nucleotide sequence conservation is shown as ribbons coloured green (indicating 90–100% conservation) and red (indicating 80–89% conservation).
https://doi.org/10.1371/journal.pgen.1011459.s004
(TIF)
S5 Fig. Gene contents of GI-pheV with different intP4 alleles from E. coli complete genomes.
Pairwise comparison of GI-pheV having the same intP4 allele, visualized by Easyfig, with black to grey gradient shows the percentage of nucleotide similarity. Strains with bolded name were compared in S4 Fig.
https://doi.org/10.1371/journal.pgen.1011459.s005
(TIF)
S1 Table. Details of the 2,382 E. coli complete genomes used in this study.
https://doi.org/10.1371/journal.pgen.1011459.s006
(XLSX)
S2 Table. Details of the reference set of 66 completely sequenced E. coli genomes used in this study.
https://doi.org/10.1371/journal.pgen.1011459.s007
(XLSX)
S3 Table. SNP distribution in gene modules and individual genes of intP4.1_GI-pheV.
https://doi.org/10.1371/journal.pgen.1011459.s008
(XLSX)
Acknowledgments
We thank Thom Cuddihy and the Queensland QFAB Bioinformatics for computational support.
References
- 1. Kaper JB, Nataro JP, Mobley HL. Pathogenic Escherichia coli. Nat Rev Microbiol. 2004;2(2):123–40. pmid:15040260.
- 2. Russo TA, Johnson JR. Proposal for a new inclusive designation for extraintestinal pathogenic isolates of Escherichia coli: ExPEC. J Infect Dis. 2000;181(5):1753–4. pmid:10823778.
- 3. Rasko DA, Rosovitz MJ, Myers GS, Mongodin EF, Fricke WF, Gajer P, et al. The pangenome structure of Escherichia coli: comparative genomic analysis of E. coli commensal and pathogenic isolates. J Bacteriol. 2008;190(20):6881–93. pmid:18676672.
- 4. Moriel DG, Tan L, Goh KG, Phan MD, Ipe DS, Lo AW, et al. A Novel Protective Vaccine Antigen from the Core Escherichia coli Genome. mSphere. 2016;1(6). pmid:27904885.
- 5. Lloyd AL, Rasko DA, Mobley HL. Defining genomic islands and uropathogen-specific genes in uropathogenic Escherichia coli. J Bacteriol. 2007;189(9):3532–46. pmid:17351047.
- 6. Petty NK, Ben Zakour NL, Stanton-Cook M, Skippington E, Totsika M, Forde BM, et al. Global dissemination of a multidrug resistant Escherichia coli clone. Proc Natl Acad Sci U S A. 2014;111(15):5694–9. pmid:24706808.
- 7. Ben Zakour NL, Alsheikh-Hussain AS, Ashcroft MM, Khanh Nhu NT, Roberts LW, Stanton-Cook M, et al. Sequential Acquisition of Virulence and Fluoroquinolone Resistance Has Shaped the Evolution of Escherichia coli ST131. mBio. 2016;7(2):e00347–16. Epub 2016/04/28. pmid:27118589.
- 8. Stoesser N, Sheppard AE, Pankhurst L, De Maio N, Moore CE, Sebra R, et al. Evolutionary History of the Global Emergence of the Escherichia coli Epidemic Clone ST131. MBio. 2016;7(2):e02162. pmid:27006459.
- 9. Decano AG, Downing T. An Escherichia coli ST131 pangenome atlas reveals population structure and evolution across 4,071 isolates. Sci Rep. 2019;9(1):17394. Epub 20191122. pmid:31758048.
- 10. Pitout JDD, Finn TJ. The evolutionary puzzle of Escherichia coli ST131. Infect Genet Evol. 2020;81:104265. Epub 20200226. pmid:32112974.
- 11. McNally A, Kallonen T, Connor C, Abudahab K, Aanensen DM, Horner C, et al. Diversification of Colonization Factors in a Multidrug-Resistant Escherichia coli Lineage Evolving under Negative Frequency-Dependent Selection. mBio. 2019;10(2). Epub 20190423. pmid:31015329.
- 12. Forde BM, Roberts LW, Phan MD, Peters KM, Fleming BA, Russell CW, et al. Population dynamics of an Escherichia coli ST131 lineage during recurrent urinary tract infection. Nat Commun. 2019;10(1):3643. Epub 20190813. pmid:31409795.
- 13. Boyd EF, Almagro-Moreno S, Parent MA. Genomic islands are dynamic, ancient integrative elements in bacterial evolution. Trends Microbiol. 2009;17(2):47–53. pmid:19162481.
- 14. Lloyd AL, Henderson TA, Vigil PD, Mobley HL. Genomic islands of uropathogenic Escherichia coli contribute to virulence. J Bacteriol. 2009;191(11):3469–81. pmid:19329634.
- 15. Hacker J, Kaper JB. Pathogenicity islands and the evolution of microbes. Annual Review Microbiology. 2000;54:641–79. pmid:11018140
- 16. Hacker J, Blum-Oehler G, Muhldorfer I, Tschape H. Pathogenicity islands of virulent bacteria: structure, function and impact on microbial evolution. Mol Microbiol. 1997;23(6):1089–97. pmid:9106201.
- 17. McInerney JO, McNally A, O’Connell MJ. Why prokaryotes have pangenomes. Nat Microbiol. 2017;2:17040. pmid:28350002.
- 18. Ingle DJ, Tauschek M, Edwards DJ, Hocking DM, Pickard DJ, Azzopardi KI, et al. Evolution of atypical enteropathogenic E. coli by repeated acquisition of LEE pathogenicity island variants. Nat Microbiol. 2016;1:15010. pmid:27571974.
- 19. Hall RJ, Whelan FJ, Cummins EA, Connor C, McNally A, McInerney JO. Gene-gene relationships in an Escherichia coli accessory genome are linked to function and mobility. Microb Genom. 2021;7(9). pmid:34499026.
- 20. Bean EL, Herman C, Anderson ME, Grossman AD. Biology and engineering of integrative and conjugative elements: Construction and analyses of hybrid ICEs reveal element functions that affect species-specific efficiencies. PLoS Genet. 2022;18(5):e1009998. Epub 20220518. pmid:35584135.
- 21. Partridge SR, Kwong SM, Firth N, Jensen SO. Mobile Genetic Elements Associated with Antimicrobial Resistance. Clin Microbiol Rev. 2018;31(4). Epub 20180801. pmid:30068738.
- 22. Welch RA, Burland V, Plunkett G 3rd, Redford P, Roesch P, Rasko D, et al. Extensive mosaic structure revealed by the complete genome sequence of uropathogenic Escherichia coli. Proc Natl Acad Sci U S A. 2002;99(26):17020–4. pmid:12471157.
- 23. Desvaux M, Dalmasso G, Beyrouthy R, Barnich N, Delmas J, Bonnet R. Pathogenicity Factors of Genomic Islands in Intestinal and Extraintestinal Escherichia coli. Front Microbiol. 2020;11:2065. Epub 20200925. pmid:33101219.
- 24. Post V, White PA, Hall RM. Evolution of AbaR-type genomic resistance islands in multiply antibiotic-resistant Acinetobacter baumannii. J Antimicrob Chemother. 2010;65(6):1162–70. pmid:20375036.
- 25. Novick RP, Schlievert P, Ruzin A. Pathogenicity and resistance islands of staphylococci. Microbes Infect. 2001;3(7):585–94. pmid:11418332.
- 26. Langille MG, Hsiao WW, Brinkman FS. Detecting genomic islands using bioinformatics approaches. Nat Rev Microbiol. 2010;8(5):373–82. pmid:20395967.
- 27. Vernikos GS, Parkhill J. Resolving the structural features of genomic islands: a machine learning approach. Genome Res. 2008;18(2):331–42. pmid:18071028.
- 28. Guo FB, Xia ZK, Wei W, Zhao HL. Statistical analyses of conserved features of genomic islands in bacteria. Genet Mol Res. 2014;13(1):1782–93. pmid:24668666.
- 29. Guyer DM, Kao JS, T MHL. Genomic analysis of a Pathogenicity Island in Uropathogenic Escherichia coli CFT073: Distribution of homologous sequences among Isolates from patients with pyelonephritis, cystitis, and catheter- associated bacteriuria and from fecal samples. Infect Immun. 1998;66(9):4411–7. pmid:9712795
- 30. Kao J-S, Stucker DM, Warren JW, LT MH. Pathogenicity island sequences of pyelonephritogenic Escherichia coli CFT073 are associated with virulent uropathogenic strains. Infect Immun. 1997;65(7):2812–20.
- 31. Rumer L, Jores J, Kirsch P, Cavignac Y, Zehmke K, Wieler LH. Dissemination of pheU- and pheV-located genomic islands among enteropathogenic (EPEC) and enterohemorrhagic (EHEC) E. coli and their possible role in the horizontal transfer of the locus of enterocyte effacement (LEE). Int J Med Microbiol. 2003;292(7–8):463–75. pmid:12635929.
- 32. Jores J, Rumer L, Kiessling S, Kaper JB, Wieler LH. A novel locus of enterocyte effacement (LEE) pathogenicity island inserted at pheV in bovine Shiga toxin-producing Escherichia coli strain O103:H2. FEMS Microbiol Lett. 2001;204(1):75–9. pmid:11682182.
- 33. Roche D, Flechard M, Lallier N, Reperant M, Bree A, Pascal G, et al. ICEEc2, a new integrative and conjugative element belonging to the pKLC102/PAGI-2 family, identified in Escherichia coli strain BEN374. J Bacteriol. 2010;192(19):5026–36. Epub 20100730. pmid:20675467.
- 34. Napolitano MG, Almagro-Moreno S, Boyd EF. Dichotomy in the evolution of pathogenicity island and bacteriophage encoded integrases from pathogenic Escherichia coli strains. Infect Genet Evol. 2011;11(2):423–36. pmid:21147268.
- 35. Morabito S, Tozzoli R, Oswald E, Caprioli A. A mosaic pathogenicity island made up of the locus of enterocyte effacement and a pathogenicity island of Escherichia coli O157:H7 is frequently present in attaching and effacing E. coli. Infect Immun. 2003;71(6):3343–8. pmid:12761117.
- 36. Hayashi T, Makino K, Ohnishi M, Kurokawa K, Ishii K, Yokoyama K, et al. Complete genome sequence of enterohemorrhagic Escherichia coli O157:H7 and genomic comparison with a laboratory strain K-12. DNA Res. 2001;8(1):11–22. pmid:11258796.
- 37. Perna NT, Plunkett G 3rd, Burland V, Mau B, Glasner JD, Rose DJ, et al. Genome sequence of enterohaemorrhagic Escherichia coli O157:H7. Nature. 2001;409(6819):529–33. pmid:11206551.
- 38. Smith EE, Sims EH, Spencer DH, Kaul R, Olson MV. Evidence for diversifying selection at the pyoverdine locus of Pseudomonas aeruginosa. J Bacteriol. 2005;187(6):2138–47. pmid:15743962.
- 39. Totsika M, Beatson SA, Sarkar S, Phan MD, Petty NK, Bachmann N, et al. Insights into a multidrug resistant Escherichia coli pathogen of the globally disseminated ST131 lineage: genome analysis and virulence mechanisms. PLoS ONE. 2011;6(10):e26578. pmid:22053197.
- 40. Owens RC Jr., Johnson JR., Stogsdill P., Yarmus L., Lolans K., Quinn J. Community transmission in the United States of a CTX-M-15-producing sequence type ST131 Escherichia coli strain resulting in death. J Clin Microbiol. 2011;49(9):3406–8. Epub 20110713. pmid:21752984.
- 41. Touchon M, Hoede C, Tenaillon O, Barbe V, Baeriswyl S, Bidet P, et al. Organised genome dynamics in the Escherichia coli species results in highly diverse adaptive paths. PLoS Genet. 2009;5(1):e1000344. Epub 20090123. pmid:19165319.
- 42. Al-Hasani K, Rajakumar K, Bulach D, Robins-Browne R, Adler B, Sakellaris H. Genetic organization of the she pathogenicity island in Shigella flexneri 2a. Microb Pathog. 2001;30(1):1–8. pmid:11162180.
- 43. Blattner FR, Plunkett G 3rd, Bloch CA, Perna NT, Burland V, Riley M, et al. The complete genome sequence of Escherichia coli K-12. Science. 1997;277(5331):1453–62. pmid:9278503.
- 44. Pierce JC, Kong D, Masker W. The effect of the length of direct repeats and the presence of palindromes on deletion between directly repeated DNA sequences in bacteriophage T7. Nucleic Acids Res. 1991;19(14):3901–5. pmid:1861982.
- 45. Albertini AM, Hofer M, Calos MP, Miller JH. On the formation of spontaneous deletions: the importance of short sequence homologies in the generation of large deletions. Cell. 1982;29(2):319–28. pmid:6288254.
- 46. Trinh TQ, Sinden RR. The influence of primary and secondary DNA structure in deletion and duplication between direct repeats in Escherichia coli. Genetics. 1993;134(2):409–22. pmid:8325478.
- 47. Bach S, Buchrieser C, Prentice M, Guiyoule A, Msadek T, Carniel E. The high-pathogenicity island of Yersinia enterocolitica Ye8081 undergoes low-frequency deletion but not precise excision, suggesting recent stabilization in the genome. Infect Immun. 1999;67(10):5091–9. pmid:10496882.
- 48. Sohanpal BK, El-Labany S, Lahooti M, Plumbridge JA, Blomfield IC. Integrated regulatory responses of fimB to N-acetylneuraminic (sialic) acid and GlcNAc in Escherichia coli K-12. Proc Natl Acad Sci U S A. 2004;101(46):16322–7. pmid:15534208.
- 49. Suzuki K, Tanabe T, Moon YH, Funahashi T, Nakao H, Narimatsu S, et al. Identification and transcriptional organization of aerobactin transport and biosynthesis cluster genes of Vibrio hollisae. Res Microbiol. 2006;157(8):730–40. pmid:16809025.
- 50. Marolda CL, Valvano MA, Crosa JH. Polymorphism in the aerobactin-cloacin DF13 receptor genes from an enteroinvasive strain of Escherichia coli and pColV-K30 is associated only with a decrease in cloacin susceptibility. Infect Immun. 1991;59(1):357–64. pmid:1987050.
- 51. Bouchet A, Valvano MA, Dho-Moulin M, Le Roy D, Andremont A. Immunological variants of the aerobactin-cloacin DF13 outer membrane protein receptor IutA among enteric bacteria. Infect Immun. 1994;62(7):3017–21. pmid:8005691.
- 52. Pitout JD, Gregson DB, Campbell L, Laupland KB. Molecular characteristics of extended-spectrum-beta-lactamase-producing Escherichia coli isolates causing bacteremia in the Calgary Health Region from 2000 to 2007: emergence of clone ST131 as a cause of community-acquired infections. Antimicrob Agents Chemother. 2009;53(7):2846–51. Epub 20090420. pmid:19380595.
- 53. Nicolas-Chanoine MH, Blanco J, Leflon-Guibout V, Demarty R, Alonso MP, Canica MM, et al. Intercontinental emergence of Escherichia coli clone O25:H4-ST131 producing CTX-M-15. J Antimicrob Chemother. 2008;61(2):273–81. Epub 20071211. pmid:18077311.
- 54. Lau SH, Kaufmann ME, Livermore DM, Woodford N, Willshaw GA, Cheasty T, et al. UK epidemic Escherichia coli strains A-E, with CTX-M-15 beta-lactamase, all belong to the international O25:H4-ST131 clone. J Antimicrob Chemother. 2008;62(6):1241–4. pmid:18779256.
- 55. Johnson JR, Tchesnokova V, Johnston B, Clabots C, Roberts PL, Billig M, et al. Abrupt emergence of a single dominant multidrug-resistant strain of Escherichia coli. J Infect Dis. 2013;207(6):919–28. pmid:23288927.
- 56. Nicolas-Chanoine MH, Bertrand X, Madec JY. Escherichia coli ST131, an intriguing clonal group. Clin Microbiol Rev. 2014;27(3):543–74. pmid:24982321.
- 57. Wirth T, Falush D, Lan R, Colles F, Mensa P, Wieler LH, et al. Sex and virulence in Escherichia coli: an evolutionary perspective. Mol Microbiol. 2006;60(5):1136–51. pmid:16689791.
- 58. Price LB, Johnson JR, Aziz M, Clabots C, Johnston B, Tchesnokova V, et al. The epidemic of extended-spectrum-beta-lactamase-producing Escherichia coli ST131 is driven by a single highly pathogenic subclone, H30-Rx. MBio. 2013;4(6):e00377–13. pmid:24345742.
- 59. Forde BM, Ben Zakour NL, Stanton-Cook M, Phan MD, Totsika M, Peters KM, et al. The complete genome sequence of Escherichia coli EC958: a high quality reference sequence for the globally disseminated multidrug resistant E. coli O25b:H4-ST131 clone. PLoS ONE. 2014;9(8):e104400. pmid:25126841.
- 60. Darling AC, Mau B, Blattner FR, Perna NT. Mauve: multiple alignment of conserved genomic sequence with rearrangements. Genome Res. 2004;14(7):1394–403. pmid:15231754.
- 61. Seemann T. Prokka: rapid prokaryotic genome annotation. Bioinformatics. 2014;30(14):2068–9. pmid:24642063.
- 62. Sullivan MJ, Petty NK, Beatson SA. Easyfig: a genome comparison visualizer. Bioinformatics. 2011;27(7):1009–10. pmid:21278367.
- 63. Carver TJ, Rutherford KM, Berriman M, Rajandream MA, Barrell BG, Parkhill J. ACT: the Artemis Comparison Tool. Bioinformatics. 2005;21(16):3422–3. pmid:15976072.
- 64. 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.
- 65. Sievers F, Wilm A, Dineen D, Gibson TJ, Karplus K, Li W, et al. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol Syst Biol. 2011;7:539. pmid:21988835.
- 66. Gouy M, Guindon S, Gascuel O. SeaView version 4: A multiplatform graphical user interface for sequence alignment and phylogenetic tree building. Mol Biol Evol. 2010;27(2):221–4. pmid:19854763.
- 67.
Synonymous Non-synonymous Analysis Program SNAP v.2.1.1. http://www.hiv.lanl.gov.
- 68.
Korber B. HIV Signature and Sequence Variation Analysis. In: Allen G. Rodrigo GHL, editor. Computational Analysis of HIV Molecular Sequences. Dordrecht, Netherlands: Kluwer Academic Publishers; 2000. p. 55–72.
- 69. Stamatakis A. RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics. 2006;22(21):2688–90. pmid:16928733.
- 70. Nhu NTK, Rahman MA, Goh KGK, Kim SJ, Phan MD, Peters KM, et al. A convergent evolutionary pathway attenuating cellulose production drives enhanced virulence of some bacteria. Nat Commun. 2024;15(1):1441. Epub 20240221. pmid:38383596.
- 71. Nhu NTK, Phan MD, Hancock SJ, Peters KM, Alvarez-Fraga L, Forde BM, et al. High-risk Escherichia coli clones that cause neonatal meningitis and association with recrudescent infection. Elife. 2024;12. Epub 20240416. pmid:38622998.
- 72. Mu A, Klare WP, Baines SL, Ignatius Pang CN, Guerillot R, Harbison-Price N, et al. Integrative omics identifies conserved and pathogen-specific responses of sepsis-causing bacteria. Nat Commun. 2023;14(1):1530. Epub 20230318. pmid:36934086.